------------------------------------------------------------------------
r4008 | vasil | 2009-01-20 17:01:08 +0200 (Tue, 20 Jan 2009) | 4 lines
branches/zip:
Add ChangeLog entries for the bugfixes in r4004 and r4005.
------------------------------------------------------------------------
r4027 | marko | 2009-01-23 11:04:49 +0200 (Fri, 23 Jan 2009) | 1 line
branches/zip: Remove some redundant #include statements.
------------------------------------------------------------------------
r4028 | marko | 2009-01-23 11:26:10 +0200 (Fri, 23 Jan 2009) | 13 lines
branches/zip: Enclose some more unused code in #ifdef UNIV_LOG_ARCHIVE.
This will help trim the dependencies of InnoDB Hot Backup.
recv_recovery_from_checkpoint_start(): Rename to
recv_recovery_from_checkpoint_start_func(), and remove the two first
parameters unless UNIV_LOG_ARCHIVE is defined. Define and use
the auxiliary macros TYPE_CHECKPOINT and LIMIT_LSN in the function.
struct recv_sys_struct: Remove archive_group unless UNIV_LOG_ARCHIVE
is defined.
Do not define LOG_ARCHIVE unless UNIV_LOG_ARCHIVE is defined.
------------------------------------------------------------------------
r4029 | marko | 2009-01-23 14:07:38 +0200 (Fri, 23 Jan 2009) | 1 line
branches/zip: Remove some redundant #include directives.
------------------------------------------------------------------------
r4030 | marko | 2009-01-23 15:31:36 +0200 (Fri, 23 Jan 2009) | 2 lines
branches/zip: buf_page_get_gen(): Remove the unused mode BUF_GET_NOWAIT.
This was noticed while investigating Issue #160.
------------------------------------------------------------------------
r4033 | marko | 2009-01-23 15:49:04 +0200 (Fri, 23 Jan 2009) | 16 lines
branches/zip: Merge revisions 4005:4032 from branches/5.1:
------------------------------------------------------------------------
r4032 | marko | 2009-01-23 15:43:51 +0200 (Fri, 23 Jan 2009) | 10 lines
branches/5.1: Merge r4031 from branches/5.0:
btr_search_drop_page_hash_when_freed(): Check if buf_page_get_gen()
returns NULL. The page may have been evicted from the buffer pool
between buf_page_peek_if_search_hashed() and buf_page_get_gen(),
because the buffer pool mutex will be released between these two calls.
(Bug #42279, Issue #160)
rb://82 approved by Heikki Tuuri
------------------------------------------------------------------------
------------------------------------------------------------------------
r4034 | marko | 2009-01-26 16:16:39 +0200 (Mon, 26 Jan 2009) | 2 lines
branches/zip: buf_page_get_gen(): Fix a "possibly uninitialized" warning
that was introduced in r4030.
------------------------------------------------------------------------
r4036 | marko | 2009-01-26 22:33:20 +0200 (Mon, 26 Jan 2009) | 22 lines
branches/zip: In r988, the function buf_block_align() was enclosed
within UNIV_DEBUG. The two remaining callers in non-debug builds,
btr_search_guess_on_hash() and btr_search_validate(), were rewritten
to call buf_page_hash_get().
To implement support for a resizeable buffer pool, the function
buf_block_align() had been rewritten to perform a page hash lookup in
the buffer pool. The caller was also made responsible for holding the
buffer pool mutex.
Because the page hash lookup is expensive and it has to be done while
holding the buffer pool mutex, implement buf_block_align() by pointer
arithmetics again, and make btr_search_guess_on_hash() call it. Note
that this will have to be adjusted if the interface to the resizeable
buffer pool is actually implemented.
rb://83 approved by Heikki Tuuri, to address Issue #161.
As a deviation from the approved patch, this patch also makes
btr_search_validate() (invoked by CHECK TABLE) check that
buf_pool->page_hash is consistent with buf_block_align().
------------------------------------------------------------------------
r4039 | vasil | 2009-01-27 08:04:17 +0200 (Tue, 27 Jan 2009) | 5 lines
branches/zip:
Adjust the paths in innodb_file_per_table.diff with a recent rename of
the test/result files in the MySQL tree.
------------------------------------------------------------------------
r4042 | marko | 2009-01-27 10:05:24 +0200 (Tue, 27 Jan 2009) | 2 lines
branches/zip: buf_LRU_invalidate_tablespace(): Fix a race condition:
read zip_size while still holding block_mutex.
------------------------------------------------------------------------
r4045 | marko | 2009-01-28 00:31:17 +0200 (Wed, 28 Jan 2009) | 8 lines
branches/zip: btr_search_validate(): Fix a bogus UNIV_DEBUG
assertion failure that was accidentally introduced in r4036.
Instead of calling buf_block_get_frame(), which asserts that the
block must be buffer-fixed, access block->frame directly. That
is safe, because changes of block->page.state are protected by
the buffer pool mutex, which we are holding.
This bug was reported by Michael.
------------------------------------------------------------------------
r4046 | marko | 2009-01-28 00:33:20 +0200 (Wed, 28 Jan 2009) | 2 lines
branches/zip: Revert the change to univ.i that was accidentally
committed in r4045.
------------------------------------------------------------------------
r4047 | marko | 2009-01-28 00:46:13 +0200 (Wed, 28 Jan 2009) | 6 lines
branches/zip: btr_search_validate(): Fix an assertion failure that was
introduced in r4036.
Do not call buf_block_get_space(), buf_block_get_page_no()
unless the block state is BUF_BLOCK_FILE_PAGE.
This bug was reported by Michael.
------------------------------------------------------------------------
r4050 | vasil | 2009-01-28 08:21:44 +0200 (Wed, 28 Jan 2009) | 5 lines
branches/zip:
Adjust the paths in innodb_lock_wait_timeout.diff with a recent rename of
the test/result files in the MySQL tree.
------------------------------------------------------------------------
r4051 | marko | 2009-01-28 14:35:49 +0200 (Wed, 28 Jan 2009) | 1 line
branches/zip: trx0sys.ic: Remove unnecessary #include <data0type.h>.
------------------------------------------------------------------------
r4052 | marko | 2009-01-28 15:21:45 +0200 (Wed, 28 Jan 2009) | 5 lines
branches/zip: Enclose some functions inside #ifdef UNIV_HOTBACKUP:
ut_sprintf_timestamp_without_extra_chars(), ut_get_year_month_day(),
log_reset_first_header_and_checkpoint(): These functions are only used
in InnoDB Hot Backup.
------------------------------------------------------------------------
r4056 | calvin | 2009-01-29 03:06:41 +0200 (Thu, 29 Jan 2009) | 33 lines
branches/zip: Merge revisions 4032:4035 from branches/5.1
All InnoDB related tests passed on Windows, except
known failure in partition_innodb_semi_consistent.
The inadvertent change to btr0sea.c in this commit is reverted in r4060.
------------------------------------------------------------------------
r4035 | vasil | 2009-01-26 09:26:25 -0600 (Mon, 26 Jan 2009) | 23 lines
branches/5.1:
Merge a change from MySQL:
------------------------------------------------------------
revno: 2646.161.4
committer: Tatiana A. Nurnberg <azundris@mysql.com>
branch nick: 51-31177v2
timestamp: Mon 2009-01-12 06:32:49 +0100
message:
Bug#31177: Server variables can't be set to their current values
Bounds-checks and blocksize corrections were applied to user-input,
but constants in the server were trusted implicitly. If these values
did not actually meet the requirements, the user could not set change
a variable, then set it back to the (wonky) factory default or maximum
by explicitly specifying it (SET <var>=<value> vs SET <var>=DEFAULT).
Now checks also apply to the server's presets. Wonky values and maxima
get corrected at startup. Consequently all non-offsetted values the user
sees are valid, and users can set the variable to that exact value if
they so desire.
------------------------------------------------------------------------
r4060 | marko | 2009-01-29 09:39:04 +0200 (Thu, 29 Jan 2009) | 1 line
branches/zip: btr0sea.c: Revert the inadvertent change made in r4056.
------------------------------------------------------------------------
------------------------------------------------------------------------
r3938 | marko | 2009-01-15 10:28:23 +0200 (Thu, 15 Jan 2009) | 3 lines
branches/zip: buf_LRU_invalidate_tablespace(), buf_LRU_free_block():
Add comments and assertions that buf_LRU_block_remove_hashed_page()
will release block_mutex when it returns BUF_BLOCK_ZIP_FREE.
------------------------------------------------------------------------
r3939 | marko | 2009-01-15 10:37:51 +0200 (Thu, 15 Jan 2009) | 7 lines
branches/zip: buf0lru.c: Improve debug assertions.
buf_LRU_block_free_non_file_page(): ut_ad(block) before dereferencing block.
buf_LRU_block_remove_hashed_page(): Forbid buf_pool_mutex_exit() while
calling buf_buddy_free(). Callers of buf_LRU_block_remove_hashed_page()
assume that the buffer pool mutex will not be released and reacquired.
------------------------------------------------------------------------
r3944 | vasil | 2009-01-15 21:15:00 +0200 (Thu, 15 Jan 2009) | 4 lines
branches/zip:
Add ChangeLog entries for the bug fixes in r3911 and r3930.
------------------------------------------------------------------------
r3958 | marko | 2009-01-16 14:53:40 +0200 (Fri, 16 Jan 2009) | 8 lines
branches/zip: Add assertions that the kernel_mutex is being held
while accessing table->locks or un_member.tab_lock.locks.
This is related to Issue #158. According to static analysis,
the added debug assertions should always hold.
lock_table_has_to_wait_in_queue(), lock_queue_iterator_reset(),
lock_queue_iterator_get_prev(), add_trx_relevant_locks_to_cache(),
fetch_data_into_cache(): Add ut_ad(mutex_own(&kernel_mutex)).
------------------------------------------------------------------------
r4006 | marko | 2009-01-20 16:29:22 +0200 (Tue, 20 Jan 2009) | 33 lines
branches/zip: Merge revisions 3930:4005 from branches/5.1:
------------------------------------------------------------------------
r4004 | marko | 2009-01-20 16:19:00 +0200 (Tue, 20 Jan 2009) | 12 lines
branches/5.1: Merge r4003 from branches/5.0:
rec_set_nth_field(): When the field already is SQL null,
do nothing when it is being changed to SQL null. (Bug #41571)
Normally, MySQL does not pass "do-nothing" updates to the storage engine.
When it does and a column of an InnoDB table that is in ROW_FORMAT=COMPACT
is being updated from NULL to NULL, the InnoDB buffer pool will be corrupted
without this fix.
rb://81 approved by Heikki Tuuri
------------------------------------------------------------------------
r4005 | marko | 2009-01-20 16:22:36 +0200 (Tue, 20 Jan 2009) | 8 lines
branches/5.1: lock_is_table_exclusive(): Acquire kernel_mutex before
accessing table->locks and release kernel_mutex before returning from
the function. This fixes a portential race condition in the
"commit every 10,000 rows" in ALTER TABLE, CREATE INDEX, DROP INDEX,
and OPTIMIZE TABLE. (Bug #42152)
rb://80 approved by Heikki Tuuri
------------------------------------------------------------------------
------------------------------------------------------------------------
r3607 | marko | 2008-12-30 22:33:31 +0200 (Tue, 30 Dec 2008) | 20 lines
branches/zip: Remove the dependency on the MySQL HASH table implementation.
Use the InnoDB hash table for keeping track of INNOBASE_SHARE objects.
struct st_innobase_share: Make table_name const uchar*. Add the member
table_name_hash.
innobase_open_tables: Change the type from HASH to hash_table_t*.
innobase_get_key(): Remove.
innobase_fold_name(): New function, for computing the fold value for the
InnoDB hash table.
get_share(), free_share(): Use the InnoDB hash functions.
innobase_end(): Free innobase_open_tables before shutting down InnoDB.
Shutting down InnoDB will invalidate all memory allocated via InnoDB.
rb://65 approved by Heikki Tuuri. This addresses Issue #104.
------------------------------------------------------------------------
r3608 | marko | 2008-12-30 22:45:04 +0200 (Tue, 30 Dec 2008) | 22 lines
branches/zip: When setting the PAGE_LEVEL of a compressed B-tree page
from or to 0, compress the page at the same time. This is necessary,
because the column information stored on the compressed page will
differ between leaf and non-leaf pages. Leaf pages are identified by
PAGE_LEVEL=0. This bug was reported as Issue #150.
Document the similarity between btr_page_create() and
btr_page_empty(). Make the function signature of btr_page_empty()
identical with btr_page_create(). (This will add the parameter "level".)
btr_root_raise_and_insert(): Replace some code with a call to
btr_page_empty().
btr_attach_half_pages(): Assert that the page level has already been
set on both block and new_block. Do not set it again.
btr_discard_only_page_on_level(): Document that this function is
probably never called. Make it work on any height tree. (Tested on
2-high tree by disabling btr_lift_page_up().)
rb://68
------------------------------------------------------------------------
r3612 | marko | 2009-01-02 11:02:44 +0200 (Fri, 02 Jan 2009) | 14 lines
branches/zip: Merge c2998 from branches/6.0, so that the same InnoDB Plugin
source tree will work both under 5.1 and 6.0. Do not add the test case
innodb_ctype_ldml.test, because it would not work under MySQL 5.1.
Refuse to create tables whose columns contain collation IDs above 255.
This removes an assertion failure that was introduced in WL#4164
(Two-byte collation IDs).
create_table_def(): Do not fail an assertion if a column contains a
charset-collation ID greater than 256. Instead, issue an error and
refuse to create the table.
The original change (branches/6.0 r2998) was rb://51 approved by Calvin Sun.
------------------------------------------------------------------------
r3613 | inaam | 2009-01-02 15:10:50 +0200 (Fri, 02 Jan 2009) | 6 lines
branches/zip: Implement the parameter innodb_use_sys_malloc
(false by default), for disabling InnoDB's internal memory allocator
and using system malloc/free instead.
rb://62 approved by Marko
------------------------------------------------------------------------
r3614 | marko | 2009-01-02 15:55:12 +0200 (Fri, 02 Jan 2009) | 1 line
branches/zip: ChangeLog: Document r3608 and r3613.
------------------------------------------------------------------------
r3615 | marko | 2009-01-02 15:57:51 +0200 (Fri, 02 Jan 2009) | 1 line
branches/zip: ChangeLog: Clarify the impact of r3608.
------------------------------------------------------------------------
r3616 | marko | 2009-01-03 00:23:30 +0200 (Sat, 03 Jan 2009) | 1 line
branches/zip: srv_suspend_mysql_thread(): Add some clarifying comments.
------------------------------------------------------------------------
r3618 | marko | 2009-01-05 12:54:53 +0200 (Mon, 05 Jan 2009) | 15 lines
branches/zip: Merge revisions 3598:3601 from branches/5.1:
------------------------------------------------------------------------
r3601 | marko | 2008-12-22 16:05:19 +0200 (Mon, 22 Dec 2008) | 9 lines
branches/5.1: Make
SET SESSION TRANSACTION ISOLATION LEVEL READ COMMITTED
a true replacement of SET GLOBAL INNODB_LOCKS_UNSAFE_FOR_BINLOG=1.
This fixes an error that was introduced in r370, causing
semi-consistent read not to not unlock rows in READ COMMITTED mode.
(Bug #41671, Issue #146)
rb://67 approved by Heikki Tuuri
------------------------------------------------------------------------
------------------------------------------------------------------------
r3623 | vasil | 2009-01-06 09:56:32 +0200 (Tue, 06 Jan 2009) | 7 lines
branches/zip:
Add patch to fix the failing main.variables mysql-test. It started failing
after the variable innodb_use_sys_malloc was added because it matches
'%alloc%' and the test is badly written and expects that no new variables
like that will ever be added.
------------------------------------------------------------------------
r3795 | marko | 2009-01-07 16:17:47 +0200 (Wed, 07 Jan 2009) | 7 lines
branches/zip: row_merge_tuple_cmp(): Do not report a duplicate key value
if any of the fields are NULL. While the tuples are equal in the
sorting order, SQL NULL is defined to be logically inequal to
anything else. (Bug #41904)
rb://70 approved by Heikki Tuuri
------------------------------------------------------------------------
r3796 | marko | 2009-01-07 16:19:32 +0200 (Wed, 07 Jan 2009) | 1 line
branches/zip: Add the tests that were forgotten from r3795.
------------------------------------------------------------------------
r3797 | marko | 2009-01-07 16:22:18 +0200 (Wed, 07 Jan 2009) | 22 lines
branches/zip: Do not call trx_allocate_for_mysql() directly, but use
helper functions that initialize some members of the transaction struct.
(Bug #41680)
innobase_trx_init(): New function: initialize some fields of a
transaction struct from a MySQL THD object.
innobase_trx_allocate(): New function: allocate and initialize a
transaction struct.
check_trx_exists(): Use the above two functions.
ha_innobase::delete_table(), ha_innobase::rename_table(),
ha_innobase::add_index(), ha_innobase::final_drop_index():
Use innobase_trx_allocate().
innobase_drop_database(): In the Windows plugin, initialize the trx_t
specially, because the THD is not available. Otherwise, use
innobase_trx_allocate().
rb://69 accepted by Heikki Tuuri
------------------------------------------------------------------------
r3798 | marko | 2009-01-07 16:42:42 +0200 (Wed, 07 Jan 2009) | 8 lines
branches/zip: row_merge_drop_temp_indexes(): Do not lock the rows of
SYS_INDEXES when looking for partially created indexes. Use the
transaction isolation level READ UNCOMMITTED to avoid interfering with
locks held by incomplete transactions that will be rolled back in a
subsequent step in the recovery. (Issue #152)
Approved by Heikki Tuuri
------------------------------------------------------------------------
r3852 | vasil | 2009-01-08 22:10:10 +0200 (Thu, 08 Jan 2009) | 4 lines
branches/zip:
Add ChangeLog entries for r3795 r3796 r3797 r3798.
------------------------------------------------------------------------
r3866 | marko | 2009-01-09 15:09:51 +0200 (Fri, 09 Jan 2009) | 2 lines
branches/zip: buf_flush_try_page(): Move some common code from each
switch case before the switch block.
------------------------------------------------------------------------
r3867 | marko | 2009-01-09 15:13:14 +0200 (Fri, 09 Jan 2009) | 2 lines
branches/zip: buf_flush_try_page(): Introduce the variable is_compressed
for caching the result of buf_page_get_state(bpage) == BUF_BLOCK_FILE_PAGE.
------------------------------------------------------------------------
r3868 | marko | 2009-01-09 15:40:11 +0200 (Fri, 09 Jan 2009) | 4 lines
branches/zip: buf_flush_insert_into_flush_list(),
buf_flush_insert_sorted_into_flush_list(): Remove unused code.
Change the parameter to buf_block_t* block and assert that
block->state == BUF_BLOCK_FILE_PAGE. This is part of Issue #155.
------------------------------------------------------------------------
r3873 | marko | 2009-01-09 22:27:40 +0200 (Fri, 09 Jan 2009) | 17 lines
branches/zip: Some non-functional changes related to Issue #155.
buf_page_struct: Note that space and offset are also protected by
buf_pool_mutex. They are only assigned to by
buf_block_set_file_page(). Thus, it suffices for buf_flush_batch() to
hold just buf_pool_mutex when checking these fields.
buf_flush_try_page(): Rename "locked" to "is_s_latched", per Heikki's request.
buf_flush_batch(): Move the common statement mutex_exit(block_mutex)
from all if-else if-else branches before the if block. Remove the
redundant test (buf_pool->init_flush[flush_type] == FALSE) that was
apparently copied from buf_flush_write_complete().
buf_flush_write_block_low(): Note why it is safe not to hold buf_pool_mutex
or block_mutex. Enumerate the assumptions in debug assertions.
------------------------------------------------------------------------
r3874 | marko | 2009-01-09 23:09:06 +0200 (Fri, 09 Jan 2009) | 4 lines
branches/zip: Add comments related to Issue #155.
buf_flush_try_page(): Note why it is safe to access bpage without
holding buf_pool_mutex or block_mutex.
------------------------------------------------------------------------
r3875 | marko | 2009-01-09 23:15:12 +0200 (Fri, 09 Jan 2009) | 11 lines
branches/zip: Non-functional change: Tighten debug assertions and
remove dead code.
buf_flush_ready_for_flush(), buf_flush_try_page(): Assert that
flush_type is one of BUF_FLUSH_LRU or BUF_FLUSH_LIST. The flush_type
comes from buf_flush_batch(), which already asserts this. The
assertion holds for all calls in the source code.
buf_flush_try_page(): Remove the dead case BUF_FLUSH_SINGLE_PAGE
of switch (flush_type).
------------------------------------------------------------------------
r3879 | marko | 2009-01-12 12:46:44 +0200 (Mon, 12 Jan 2009) | 14 lines
branches/zip: Simplify the flushing of dirty pages from the buffer pool.
buf_flush_try_page(): Rename to buf_flush_page(), and change the
return type to void. Replace the parameters space, offset with bpage,
and remove the second page hash lookup. Note and assert that both
buf_pool_mutex and block_mutex must now be held upon entering the
function. They will still be released by this function.
buf_flush_try_neighbors(): Replace buf_flush_try_page() with
buf_flush_page(). Make the logic easier to follow by not negating the
precondition of buf_flush_page().
rb://73 approved by Sunny Bains. This is related to Issue #157.
------------------------------------------------------------------------
r3880 | marko | 2009-01-12 13:24:37 +0200 (Mon, 12 Jan 2009) | 2 lines
branches/zip: buf_flush_page(): Fix a comment that should have been fixed
in r3879. Spotted by Sunny.
------------------------------------------------------------------------
r3881 | marko | 2009-01-12 14:25:22 +0200 (Mon, 12 Jan 2009) | 2 lines
branches/zip: buf_page_get_newest_modification(): Use the block mutex
instead of the buffer pool mutex. This is related to Issue #157.
------------------------------------------------------------------------
r3882 | marko | 2009-01-12 14:40:08 +0200 (Mon, 12 Jan 2009) | 3 lines
branches/zip: struct mtr_struct: Remove the unused field magic_n
unless UNIV_DEBUG is defined. mtr->magic_n is only assigned to
and checked in UNIV_DEBUG builds.
------------------------------------------------------------------------
r3883 | marko | 2009-01-12 14:48:59 +0200 (Mon, 12 Jan 2009) | 1 line
branches/zip: Non-functional change: Use ut_d when assigning to mtr->state.
------------------------------------------------------------------------
r3884 | marko | 2009-01-12 18:56:11 +0200 (Mon, 12 Jan 2009) | 16 lines
branches/zip: Non-functional change: Add some debug assertions and comments.
buf_page_t: Note that the LRU fields are protected by buf_pool_mutex
only, not block->mutex or buf_pool_zip_mutex.
buf_page_get_freed_page_clock(): Note that this is sometimes invoked
without mutex protection.
buf_pool_get_oldest_modification(): Note that the result may be out of
date.
buf_page_get_LRU_position(), buf_page_is_old(): Assert that the buffer
pool mutex is being held.
buf_page_release(): Assert that dirty blocks are in the flush list.
------------------------------------------------------------------------
r3896 | marko | 2009-01-13 09:30:26 +0200 (Tue, 13 Jan 2009) | 2 lines
branches/zip: buf_flush_try_neighbors(): Fix a bug
that was introduced in r3879 (rb://73).
------------------------------------------------------------------------
r3900 | marko | 2009-01-13 10:32:24 +0200 (Tue, 13 Jan 2009) | 1 line
branches/zip: Fix some comments to say buf_pool_mutex.
------------------------------------------------------------------------
r3907 | marko | 2009-01-13 11:54:01 +0200 (Tue, 13 Jan 2009) | 3 lines
branches/zip: row_merge_create_temporary_table(): On error,
row_create_table_for_mysql() already frees new_table.
Do not attempt to free it again.
------------------------------------------------------------------------
r3908 | marko | 2009-01-13 12:34:32 +0200 (Tue, 13 Jan 2009) | 1 line
branches/zip: Enable HASH_ASSERT_OWNED independently of UNIV_SYNC_DEBUG.
------------------------------------------------------------------------
r3914 | marko | 2009-01-13 21:46:22 +0200 (Tue, 13 Jan 2009) | 37 lines
branches/zip: In hash table lookups, assert that the traversed items
satisfy some conditions when UNIV_DEBUG is defined.
HASH_SEARCH(): New parameter: ASSERTION. All users will pass an appropriate
ut_ad() or nothing.
dict_table_add_to_columns(): Assert that the table being added to the data
dictionary cache is not already being pointed to by the name_hash and
id_hash tables.
HASH_SEARCH_ALL(): New macro, for use in dict_table_add_to_columns().
dict_mem_table_free(): Set ut_d(table->cached = FALSE), so that we can
check ut_ad(table->cached) when traversing the hash tables, as in
HASH_SEARCH(name_hash, dict_sys->table_hash, ...) and
HASH_SEARCH(id_hash, dict_sys->table_id_hash, ...).
dict_table_get_low(), dict_table_get_on_id_low(): Assert
ut_ad(!table || table->cached).
fil_space_get_by_id(): Check ut_ad(space->magic_n == FIL_SPACE_MAGIC_N)
in HASH_SEARCH(hash, fil_system->spaces, ...).
fil_space_get_by_name(): Check ut_ad(space->magic_n == FIL_SPACE_MAGIC_N)
in HASH_SEARCH(name_hash, fil_system->name_hash, ...).
buf_buddy_block_free(): Check that the blocks are in valid state in
HASH_SEARCH(hash, buf_pool->zip_hash, ...).
buf_page_hash_get(): Check that the blocks are in valid state in
HASH_SEARCH(hash, buf_pool->page_hash, ...).
get_share(), free_share(): Check ut_ad(share->use_count > 0) in
HASH_SEARCH(table_name_hash, innobase_open_tables, ...).
This was posted as rb://75 for tracking down errors similar to Issue #153.
------------------------------------------------------------------------
r3931 | marko | 2009-01-14 16:06:22 +0200 (Wed, 14 Jan 2009) | 26 lines
branches/zip: Merge revisions 3601:3930 from branches/5.1:
------------------------------------------------------------------------
r3911 | sunny | 2009-01-13 14:15:24 +0200 (Tue, 13 Jan 2009) | 13 lines
branches/5.1: Fix Bug#38187 Error 153 when creating savepoints
InnoDB previously treated savepoints as a stack e.g.,
SAVEPOINT a;
SAVEPOINT b;
SAVEPOINT c;
SAVEPOINT b; <- This would delete b and c.
This fix changes the behavior to:
SAVEPOINT a;
SAVEPOINT b;
SAVEPOINT c;
SAVEPOINT b; <- Does not delete savepoint c
------------------------------------------------------------------------
r3930 | marko | 2009-01-14 15:51:30 +0200 (Wed, 14 Jan 2009) | 4 lines
branches/5.1: dict_load_table(): If dict_load_indexes() fails,
invoke dict_table_remove_from_cache() instead of dict_mem_table_free(),
so that the data dictionary will not point to freed data.
(Bug #42075, Issue #153, rb://76 approved by Heikki Tuuri)
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
r3572 | marko | 2008-12-17 11:19:56 +0200 (Wed, 17 Dec 2008) | 3 lines
Changed paths:
M /branches/zip/ChangeLog
M /branches/zip/btr/btr0sea.c
M /branches/zip/buf/buf0buf.c
M /branches/zip/buf/buf0lru.c
M /branches/zip/ha/ha0ha.c
M /branches/zip/ha/hash0hash.c
M /branches/zip/include/buf0buf.h
M /branches/zip/include/ha0ha.h
M /branches/zip/include/ha0ha.ic
M /branches/zip/include/hash0hash.h
M /branches/zip/include/univ.i
branches/zip: Introduce UNIV_AHI_DEBUG for debugging the adaptive hash
index without enabling UNIV_DEBUG.
------------------------------------------------------------------------
r3574 | marko | 2008-12-17 12:44:31 +0200 (Wed, 17 Dec 2008) | 2 lines
Changed paths:
M /branches/zip/ChangeLog
branches/zip: ChangeLog: Document recent changes that were not included in
InnoDB Plugin 1.0.2, except changes to source code comments.
------------------------------------------------------------------------
r3575 | marko | 2008-12-17 14:40:58 +0200 (Wed, 17 Dec 2008) | 12 lines
Changed paths:
M /branches/zip/ChangeLog
M /branches/zip/include/row0sel.h
M /branches/zip/include/row0upd.h
M /branches/zip/pars/pars0pars.c
M /branches/zip/row/row0mysql.c
M /branches/zip/row/row0sel.c
M /branches/zip/row/row0upd.c
branches/zip: Remove update-in-place-in-select from the internal SQL
interpreter. It was only used for updating the InnoDB internal data
dictionary when renaming or dropping tables. It could have caused
deadlocks after acquiring latches on insert buffer bitmap pages.
This and r3544 should fix Issue #135.
Furthermore, the update-in-place-in-select does not account for
compression failure. That was not a problem yet, since the InnoDB SQL
interpreter has so far assumed ROW_FORMAT=REDUNDANT.
rb://63 approved by Heikki Tuuri
------------------------------------------------------------------------
------------------------------------------------------------------------
r3537 | marko | 2008-12-16 10:24:03 +0200 (Tue, 16 Dec 2008) | 3 lines
branches/zip: sync_thread_add_level(): Add a comment explaining the
assertion about SYNC_BUF_POOL and SYNC_BUF_BLOCK.
------------------------------------------------------------------------
r3540 | marko | 2008-12-16 12:13:31 +0200 (Tue, 16 Dec 2008) | 2 lines
branches/zip: buf_page_init_for_read(): Use common code for error exit.
------------------------------------------------------------------------
r3541 | marko | 2008-12-16 12:14:58 +0200 (Tue, 16 Dec 2008) | 3 lines
branches/zip: btr_cur_optimistic_delete(): Note that no further pages
must be latched before calling mtr_commit(mtr) if the function returns TRUE.
------------------------------------------------------------------------
watch. This should fix the race condition that seems to have caused
Issue #126: When another thread is buffering an insert for the record
that is being purged, it should invoke buf_pool_watch_notify(), so
that the purge will not be buffered. Otherwise, the purge would be
buffered for the wrong record (one that was inserted after the purge
determined that the record can be removed).
When deletes are not buffered, the latch on the secondary index page
would prevent the insert from occurring. In delete buffering, the
buf_pool_watch replaces the page latch when the page is not in the
buffer pool.
buf_pool_watch_notify(): Make public.
ibuf_insert(): Invoke buf_pool_watch_notify() when buffering an
insert or a delete-mark operation.
This is to fix issues introduced by recovery enhancements committed
through r2993. The fix is to update flush_rbt whenever flush_list is
manipulated because of buf_page/buf_block relocation.
rb://54
Reviewed by: Marko
------------------------------------------------------------------------
r3180 | marko | 2008-11-21 16:36:18 +0200 (Fri, 21 Nov 2008) | 3 lines
branches/zip: buf_LRU_free_block(): Do not assert on ibuf_count until after
it has been determined that the block is not being I/O-fixed.
This corrects the mistake that was made in r3177.
------------------------------------------------------------------------
------------------------------------------------------------------------
r3170 | marko | 2008-11-21 10:11:18 +0200 (Fri, 21 Nov 2008) | 4 lines
branches/zip: dtuple_print(): Dump each field in a separate line,
so that the dumps can be read and compared more easily.
This is related to the change to rec_print_old() and rec_print_comp()
in r3148.
------------------------------------------------------------------------
r3177 | marko | 2008-11-21 16:24:31 +0200 (Fri, 21 Nov 2008) | 3 lines
branches/zip: buf_LRU_free_block(), buf_page_try_get_func(): Assert that
there is nothing in the insert buffer for the page. This is for
tracking down Issue #128.
------------------------------------------------------------------------
------------------------------------------------------------------------
r3036 | marko | 2008-11-12 12:34:30 +0200 (Wed, 12 Nov 2008) | 4 lines
branches/zip: dtuple_validate(): When UNIV_DEBUG_VALGRIND is defined,
rely solely on the UNIV_MEM_ASSERT_RW() check and disable the for loop
that would only cause additional noise.
------------------------------------------------------------------------
r3037 | marko | 2008-11-12 13:52:57 +0200 (Wed, 12 Nov 2008) | 6 lines
branches/zip: row_vers_impl_x_locked_off_kernel(): Remove compilation
warnings about prev_trx_id and vers_del being possibly uninitialized,
by handling the case prev_version == NULL in a single if block.
rb://45 approved by Inaam Rana.
------------------------------------------------------------------------
r3131 | michael | 2008-11-17 14:56:56 +0200 (Mon, 17 Nov 2008) | 9 lines
branches/zip:
rb://53
Improve innodb_supports_xa system variable handling and
reduces the number of retrievals of the value from MySQL.
Approved by: Marko, over IM
------------------------------------------------------------------------
r3132 | michael | 2008-11-17 16:02:01 +0200 (Mon, 17 Nov 2008) | 5 lines
branches/zip: rb://53
Final version of rb://53, fixes the styling of a comment, makes
the definition and the declaration of thd_supports_xa() identical commentwise.
------------------------------------------------------------------------
r3141 | marko | 2008-11-19 16:39:55 +0200 (Wed, 19 Nov 2008) | 1 line
branches/zip: buf_LRU_free_block(): Clarify the function comment.
------------------------------------------------------------------------
r3144 | marko | 2008-11-20 11:39:49 +0200 (Thu, 20 Nov 2008) | 2 lines
branches/zip: rec_get_nth_field_offs_old(): Add UNIV_UNLIKELY hints
to assertion-like tests.
------------------------------------------------------------------------
r3145 | marko | 2008-11-20 12:22:40 +0200 (Thu, 20 Nov 2008) | 20 lines
branches/zip: Always check for "row too large" when executing SQL to create
an index or table. We have to skip this check when loading table definitions
from the data dictionary, because we could otherwise refuse to load old
tables (even uncompressed ones). This addresses Issue #119.
The first "row too large" check was implemented in MySQL 5.0.3
to address MySQL Bug #5682. In the InnoDB Plugin 1.0.2, a more
accurate check was implemented in innodb_strict_mode. We now
make the check unconditional.
dict_create_index_step(): Pass strict=TRUE to dict_index_add_to_cache().
trx_is_strict(), thd_is_strict(): Remove.
innodb-zip.test: Test in innodb_strict_mode=OFF.
innodb_bug36169.test: Ensure that none of the tables can be created.
rb://56 approved by Sunny Bains.
------------------------------------------------------------------------
r3148 | marko | 2008-11-20 13:27:27 +0200 (Thu, 20 Nov 2008) | 3 lines
branches/zip: rec_print_old(), rec_print_comp(): Dump each field in a
separate line, so that the dumps can be read and compared more easily.
------------------------------------------------------------------------
This patch is to improve recovery performance in InnoDB+.
It includes introduction of red-black tree for sorted insertion into
the flush_list and couple of other quirks. More can be found
at: https://svn.innodb.com/innobase/Recovery_Performance_Improvements
Reviewed by: Marko
that were accidentally committed in r2971:
univ.i, ha_innobase::create(): Reduce the uncompressed page size from
16 to 4 kilobytes.
btr_cur_search_to_nth_level(): Print a diagnostic message when the
leaf page of a secondary index is being requested from the disk.
buf_read_ahead_random(), buf_read_ahead_linear(): Disable.
Invoke ibuf_get_volume_buffered_count() as many times as
ibuf_rec_get_volume(rec), so that *n_recs will be updated as appropriate.
This fixes Issue #82: some deletes will be buffered in purge.
buf_pool->watch_ fields to start with the common prefix buf_pool_watch,
so that they can be searched easily:
buf_pool_watch_set(): Renamed from buf_pool_add_watch(). The "add" was
misleading, because only one watch can be active at a time.
buf_pool_watch_clear(): Renamed from buf_pool_remove_watch().
buf_pool_watch_occurred(): Renamed from buf_pool_watch_happened().
buf_page_get_gen(): Once again, zip_size must always match the
compressed page size of the tablespace where the page is requested from.
This seems to hold also for all calls from the insert buffer.
buf_page_optimistic_get_func(): Do not pass mtr to ibuf_page(), because
there is no guarantee that mtr would contain an x-latch to the insert
buffer bitmap page that covers (block->page.space, block->page.offset).
zip_size should always equal fil_space_get_zip_size(space), even when called
from ibuf.
ibuf_page() should be called with mtr=NULL to be equivalent to the original
implementation. We cannot assume that mtr holds an x-latch on the insert
buffer bitmap page that covers the page that is being requested.
from branches/zip.
btr_cur_optimistic_delete(): Split a too long line in the way it is
split in branches/zip.
buf_page_get_gen(): Add missing space in the function comment.
buf_print_io(): Restoer a removed space to the printout.
univ.i: Remove C++-style (or C99-style) comments.
buf0buf.h: Undo the white-space changes to the wrapper macros of
buf_page_get_gen().
ibuf_update_free_bits_low(): Revert the changes. Restore the assertion
that this function must not be invoked on compressed pages. The function
ibuf_update_free_bits_zip() is for compressed pages.
ibuf_insert_to_index_page(): Undo a white-space change.
The followin mysql-tests failed (and they are known to fail):
main.information_schema [ fail ]
main.innodb_file_per_table_basic[ fail ]
main.type_bit_innodb [ fail ]
Tested against : MYSQL_SERVER_VERSION "5.1.28"
buf_LRU_old_adjust_len(): Replace a constant ut_ad() with a preprocessor check.
buf_LRU_free_block(): Remove the check for the unlikely case
buf_pool->LRU_old == prev_b in order to simplify the function.
The check was implemented as part of r2306.
symbols. Use it for all definitions of non-static variables and functions.
lexyy.c, make_flex.sh: Declare yylex as UNIV_INTERN, not static. It is
referenced from pars0grm.c.
Actually, according to
nm .libs/ha_innodb.so|grep -w '[ABCE-TVXYZ]'
the following symbols are still global:
* The vtable for class ha_innodb
* pars0grm.c: The function yyparse() and the variables yychar, yylval, yynerrs
The required changes to the Bison-generated file pars0grm.c will be addressed
in a separate commit, which will add a script similar to make_flex.sh.
The class ha_innodb is renamed from class ha_innobase by a #define. Thus,
there will be no clash with the builtin InnoDB. However, there will be some
overhead for invoking virtual methods of class ha_innodb. Ideas for making
the vtable hidden are welcome. -fvisibility=hidden is not available in GCC 3.
stored columns (BLOBs).
btr_copy_blob_prefix(), btr_copy_zblob_prefix(),
btr_copy_externally_stored_field_prefix_low(),
btr_copy_externally_stored_field_prefix(),
btr_copy_externally_stored_field(),
btr_rec_copy_externally_stored_field():
Note that the page containing the clustered index record that points to
the BLOB must be latched.
btr_copy_zblob_prefix(): Note that there is no latch on the page, and thus
all accesses to a given page via this function must be covered by the same
set of locks or latches.
btr_copy_zblob_prefix(): Note that the block acquired by
buf_page_get_zip() is protected by an exclusive table lock or
or by a latch on the clustered index record.
of the buffer pool mutex. Apparently, it is temporarily released also
in older versions of MySQL/InnoDB for the duration of the
btr_search_drop_page_hash_index() call [in buf_LRU_free_block()].
the buffer pool mutex. The instrumentation can be activated by
defining UNIV_DEBUG or UNIV_BUF_DEBUG at compilation time.
buf_pool_mutex_exit_forbidden: New variable. When this is nonzero,
an assertion will fail in buf_pool_mutex_exit().
buf_pool_mutex_exit_forbid(): Macro for declaring that the buffer pool
mutex must not be released. Calls may be nested.
buf_pool_mutex_exit_allow(): Macro for declaring that the buffer pool
mutex may be released. Calls may be nested.
buf_LRU_search_and_free_block(): Prohibit buf_pool_mutex_exit() in the
scope of the function.
buf_LRU_free_block(): Prohibit buf_pool_mutex_exit() in buf_buddy_alloc()
and buf_buddy_free().
buf_LRU_block_remove_hashed_page(): Prohibit buf_pool_mutex_exit()
in buf_buddy_free().
buf_pool->mutex: Rename to buf_pool_mutex, so that the wrappers will have
to be used when changes are merged from other source trees.
buf_pool->zip_mutex: Rename to buf_pool_zip_mutex.
buf_pool_mutex_own(), buf_pool_mutex_enter(), buf_pool_mutex_exit():
Wrappers for buf_pool_mutex.
buf_buddy_min_n_frames into two, also buf_buddy_max_n_frames.
Set the default values in such a way that a strict LRU policy will apply
for replacing compressed or uncompressed page frames in the buffer pool.
These parameters have not yet been exposed to the MySQL layer.
allocating compressed page frames or their control blocks. Also note
that if buf_buddy_alloc() is used for allocating a control block,
it must be initialized before releasing buf_pool->mutex.
buf_page_init_for_read(): When the page hash check fails after
buf_buddy_alloc(), free the uninitialized control block before freeing
the compressed page frame. This fixes a potential error in
buf_buddy_relocate_block().
mutex is temporarily released.
buf_LRU_free_block(), buf_buddy_alloc_clean(): Add an output parameter that
will be assigned TRUE when the buffer pool mutex is released.
This bug was spotted by and fix provided by Sunny.
