- 30 Oct, 2006 1 commit
-
-
marko authored
order to be violated if the thread holds dict_operation_lock, whose level is SYNC_DICT_OPERATION. This removes the assertion failure of TRUNCATE TABLE #ifdef UNIV_SYNC_DEBUG.
-
- 26 Oct, 2006 1 commit
-
-
marko authored
ChangeSet 2006/10/09 19:51:41+04:00 gkodinov@macbook.local Bug #22781: SQL_BIG_RESULT fails to influence sort plan Currently SQL_BIG_RESULT is checked only at compile time. However, additional optimizations may take place after this check that change the sort method from 'filesort' to sorting via index. As a result the actual plan executed is not the one specified by the SQL_BIG_RESULT hint. Similarly, there is no such test when executing EXPLAIN, resulting in incorrect output. The patch corrects the problem by testing for SQL_BIG_RESULT both during the explain and execution phases.
-
- 19 Oct, 2006 1 commit
-
-
osku authored
in the father page. btr0btr.h: Add a BTR_MAX_LEVELS define.
-
- 18 Oct, 2006 1 commit
-
-
marko authored
instead of passing the wrong length to ut_print_namel(). (Bug #23368) ut_print_name(), ut_print_namel(): Document the special treatment of '/'. ut_print_namel(): Replace strchr() with memchr(), as the string might not be NUL-terminated.
-
- 16 Oct, 2006 2 commits
- 12 Oct, 2006 1 commit
-
-
marko authored
UNIV_SYNC_DEBUG is defined. rw_lock_list_print_info(): Add parameter "file". mutex_list_print_info(): Add parameter "file". Make the function static.
-
- 10 Oct, 2006 2 commits
- 09 Oct, 2006 3 commits
-
-
marko authored
Replace the search loop with a call to dict_index_find_on_id_low().
-
marko authored
managed with an LRU algorithm.
-
marko authored
ha_innodb.cc: innodb_mutex_show_status(): Add (ulong) casts to ulonglong expressions being passed to fprintf %lu. The warnings were apparently introduced by MySQL AB developers. mysql_declare_plugin(innobase): Add PLUGIN_LICENSE_GPL. have_innodb.inc: Merge changes from MySQL AB.
-
- 06 Oct, 2006 1 commit
-
-
marko authored
in the "#else" branch of "#if building on Windows".
-
- 05 Oct, 2006 2 commits
-
-
marko authored
dict_col_copy_type_noninline(): Remove the return statement. Some compilers do not allow return expressions of void type. innobase_start_or_create_for_mysql(): Do not cap srv_max_threads to 1000 on Windows. Makefile.am (EXTRA_DIST): Add plug.in
-
marko authored
innodb_unsafe_binlog, innodb_handler, innodb-big, innodb_cache-master, innodb-deadlock, innodb_cache: Remove. These have been renamed in the MySQL tree. innodb_mysql: Update.
-
- 03 Oct, 2006 3 commits
-
-
marko authored
ChangeSet 2006/09/30 18:44:42-07:00 brian@zim.(none) Merge zim.(none):/home/brian/mysql/merge-5.1 into zim.(none):/home/brian/mysql/arch-5.1 ChangeSet 2006/09/30 12:49:46-07:00 brian@zim.(none) This patch adds handlerton passing to functions. NDB and Innodb still require a global hanlderton in the main code due to the nature of the sql_cache call back function (should be solveable... another patch). Partitioning now has a flag to allow disabling of engines from being compatible with partitioning. Cleaned up heap naming convention on panic call. ChangeSet 2006/09/29 17:19:02-07:00 brian@zim.(none) This removes the passing of global hton to engine instance. ChangeSet 2006/09/28 13:22:56+02:00 gbichot@dl145h.mysql.com Merge gbichot@bk-internal:/home/bk/mysql-5.1-arch into dl145h.mysql.com:/users/gbichot/mysql-5.1-arch ChangeSet 2006/09/28 13:19:43+02:00 gbichot@dl145h.mysql.com In the handlerton, cursor creation function don't have an argument and so the engine calls current_thd to derive transaction information; instead we now pass THD to those functions, it looks more logical (it makes the implicit current_thd parameter more visible). Approved by Brian and Monty. ChangeSet 2006/09/26 22:51:53-07:00 brian@zim.(none) Merge zim.(none):/home/brian/mysql/merge-5.1 into zim.(none):/home/brian/mysql/arch-5.1
-
marko authored
the current directory.
-
marko authored
-
- 02 Oct, 2006 1 commit
-
-
marko authored
-
- 26 Sep, 2006 2 commits
-
-
marko authored
bug would only surface if the InnoDB SQL interpreter were used for updating fixed-length columns. Currently (as the UPD_NODE_NO_SIZE_CHANGE flag is never set), cmpl_info can only be 0 or UPD_NODE_NO_ORD_CHANGE. Luckily, UPD_NODE_NO_ORD_CHANGE is 1, and the condition was simplified to !cmpl_info.
-
marko authored
-
- 22 Sep, 2006 1 commit
-
-
marko authored
to comply with ISO/IEC 9899:1989. This error was made in r834. (Bug #22575)
-
- 21 Sep, 2006 1 commit
-
-
osku authored
tree.
-
- 20 Sep, 2006 5 commits
- 19 Sep, 2006 3 commits
- 18 Sep, 2006 6 commits
-
-
marko authored
-
marko authored
so that all integer fields can be packed into 64 bits. (Bug #20877) dtype_t: Change the type of all bit-fields to unsigned. dict_table_get_nth_col(), dict_table_get_sys_col_noninline(), dict_table_get_sys_col(), dict_field_get_col(): Return const dict_col_t*, so that changes to dict_col_t can be detected more easily. Add const to many dict_col_t* declarations. dict_index_get_nth_type(): Replace with dict_index_get_nth_col(). dict_col_get_type(): Replace with dict_col_copy_type(). dict_col_get_min_size(), dict_col_get_max_size(), dict_col_get_fixed_size(), dict_col_get_sql_null_size(): New functions. dtype_get_at_most_n_mbchars(): Replace the parameter dtype with the parameters prtype, mbminlen, mbmaxlen. dtype_get_pad_char(), cmp_data_data(), cmp_data_data_slow(), cmp_whole_field(): Replace the dtype_t* parameter with the ulint parameters mtype, prtype. dtype_copy(): Add a const qualifier to type2 (the one being copied from). dtype_set_mblen(): Replaced with dtype_get_mblen(). dtype_get_fixed_size_low(), dtype_get_min_size_low(), dtype_get_fixed_max_low(): Replace dtype_get_fixed_size(), dtype_get_min_size(), and dtype_get_max_size(). These are used by the dict_col_get_{fixed,min,max}_size() functions. cmp_types_are_equal(): Replace with cmp_cols_are_equal(). dict_table_get_col_name(): Add a const qualifier parameter to the parameter "table". dtype_binary, dtype_binary_val: Remove. dtype_is_fixed_size(): Remove.
-
marko authored
rw_lock_t: Do not make writer_is_wait_ex a bit-field. There are no fields yet that could be fused to the same machine word, but we play it safe, because the field writer_is_wait_ex may be modified by several threads simultaneously. Such fields should always be allocated an own machine word. dict_table_t: Change the type of all bit-fields to "unsigned". Make "space" a bit-field of 32 bits. Move name_hash and id_hash after all bit-fields, so that the bit-fields can be allocated together. Do not make autoinc_inited a bit-field, as we cannot allocate any field from the same machine word. dict_build_table_def_step(): Pass a local variable to fil_create_new_single_table_tablespace() and initialize table->space from it after the call, now that table->space is a bit-field.
-
osku authored
dict_col_t::name, and use it instead of col->name everywhere.
-
marko authored
-
marko authored
ut_align_down(ptr, UNIV_PAGE_SIZE) calls.
-
- 15 Sep, 2006 3 commits
-
-
heikki authored
it should be redo-logged because the data structure is file-based; this patch does not fix any bug; the original erase operation was added in r781 to fix Valgrind Bug #20791
-
marko authored
There always was a one-to-one mapping between dict_tree_t and dict_index_t. This saves 6 machine words per B-tree index in the data dictionary cache plus the memory allocation overhead. We save one mem_heap_t object per index (15 machine words). Considering the internal fragmentation of the buddy allocator in mem_area_alloc(), this should save 32 machine words per index (128 bytes on 32-bit systems and 256 bytes on 64-bit systems). (Bug #20877) struct dict_tree_struct, dict_tree_t: Remove. struct dict_index_struct: Add page and lock. dict_tree_create(): Remove. Replace the invocation with assignment to index->page and a call to rw_lock_create(&index->lock). dict_tree_free(): Remove. Replace the invocation wtih a call to rw_lock_free(&index->lock). dict_index_get_tree(): Remove. dict_tree_get_space_reserve(): Rename to dict_index_get_space_reserve() and remove the parameter, which was unused. btr_level_list_remove(): Remove the unused parameter "tree". Replace the occurrences of "tree" with "index" in names of variables, functions and data types, e.g. "dict_tree_t tree" becomes "dict_index_t index". Remove local variables "tree" or "index" of functions that needed both "tree" and "index".
-
osku authored
-