~stewart/haildb/embedded-innodb-1.0.6.6750-branch-point

• Changes
• Files

~stewart/haildb/embedded-innodb-1.0.6.6750-branch-point » Changes from revision 2028

From Revision 2007 to 1988

expand all

collapse all

Rev		Summary	Authors	Date	Diff	Files
2007		branches/zip: trx_rseg_create(): Unused function, remove. branches/zip: trx_rseg_create(): Unused function, remove.	marko	14 years ago
2006		branches/zip: Clean up after r6559. Now that branches/zip: Clean up after r6559.  Now that btr_pcur_open_with_no_init() is a macro, do not mix preprocessor directives in the macro invocation, because it is implementation-defined whether that is going to work.	marko	14 years ago
2005		branches/zip: Merge revisions 6538:6613 from branches/5.1: branches/zip: Merge revisions 6538:6613 from branches/5.1:   ------------------------------------------------------------------------   r6545 | jyang | 2010-02-03 03:57:32 +0200 (Wed, 03 Feb 2010) | 8 lines   Changed paths:      M /branches/5.1/lock/lock0lock.c   branches/5.1: Fix bug #49001, "SHOW INNODB STATUS deadlock info   incorrect when deadlock detection aborts". Print the correct   lock owner when recursive function lock_deadlock_recursive()   exceeds its maximum depth LOCK_MAX_DEPTH_IN_DEADLOCK_CHECK.   rb://217, approved by Marko.   ------------------------------------------------------------------------   r6613 | inaam | 2010-02-09 20:23:09 +0200 (Tue, 09 Feb 2010) | 11 lines   Changed paths:      M /branches/5.1/buf/buf0buf.c      M /branches/5.1/buf/buf0rea.c      M /branches/5.1/include/buf0rea.h   branches/5.1: Fix Bug #38901   InnoDB logs error repeatedly when trying to load page into buffer pool   In buf_page_get_gen() if we are unable to read a page (because of   corruption or some other reason) we keep on retrying. This fills up   error log with millions of entries in no time and we'd eventually run   out of disk space. This patch limits the number of attempts that we   make (currently set to 100) and after that we abort with a message.   rb://241 Approved by: Heikki   ------------------------------------------------------------------------	marko	14 years ago
2004		branches/zip: Document r6614 in ChangeLog. branches/zip: Document r6614 in ChangeLog.	marko	14 years ago
2003		branches/plugin rb://242 branches/plugin rb://242 Let the master thread sleep if the amount of work to be done is calibrated as taking less than a second. Approved by: Heikki	inaam	14 years ago
2002		branches/zip: recv_recovery_rollback_active(): branches/zip: recv_recovery_rollback_active(): Drop the temporary tables and indexes after enabling sync order checks. This should not make any difference. This could have been done in r6611.	marko	14 years ago
2001		branches/zip: Roll back dictionary transaction(s) before sca... branches/zip: Roll back dictionary transaction(s) before scanning *.ibd files innobase_start_or_create_for_mysql(): Roll back data dictionary transactions before scanning the *.ibd files. Then, data dictionary records can be loaded to the cache before opening the *.ibd files. recv_recovery_rollback_active(): Refactored from recv_recovery_from_checkpoint_finish(). rb://235, committing without review, because this is needed for TablespaceDictionary.	marko	14 years ago
2000		branches/zip: When dropping temporary indexes and tables at ... branches/zip: When dropping temporary indexes and tables at startup, first load them to the data dictionary cache and use the normal routines for dropping tables or indexes.  This should reduce the risk of bugs and also make the code compatible with the upcoming TablespaceDictionary implementation. DICT_SYS_INDEXES_NAME_FIELD: The clustered index position of SYS_INDEXES.NAME. row_merge_drop_temp_indexes(): Scan SYS_INDEXES for tables containing temporary indexes, and load the tables as needed. Invoke row_merge_drop_index() to drop the indexes. row_mysql_drop_temp_tables(): Scan SYS_TABLES for temporary tables, load them with dict_load_table() and drop them with row_drop_table_for_mysql(). rb://251, not yet reviewed	marko	14 years ago
1999		branches/zip: dict_field_print_low(): Add const qualifier. branches/zip: dict_field_print_low(): Add const qualifier.	marko	14 years ago
1998		branches/zip: ha_innobase::add_index(): Check for !innodb_ta... branches/zip: ha_innobase::add_index(): Check for !innodb_table.	marko	14 years ago
1997		branches/zip: btr_pcur_commit(): Unused function, remove. branches/zip: btr_pcur_commit(): Unused function, remove.	marko	14 years ago
1996		branches/zip: rec_get_nth_field_offs_old(): branches/zip: rec_get_nth_field_offs_old(): Replace if (!cond) ut_error; tests with ut_a(cond).	marko	14 years ago
1995		branches/zip: row_merge_drop_index(): Remove redundant condi... branches/zip: row_merge_drop_index(): Remove redundant condition on SYS_INDEXES.TABLE_ID.  INDEX_ID must be instance-widely unique, because SYS_FIELDS is not indexed by TABLE_ID.	marko	14 years ago
1994		branches/zip: Remove the additional check introduced in r653... branches/zip: Remove the additional check introduced in r6534 which tries to check if the joining transaction has any other transactions waiting on its locks. This optimization results in excessive deadlocks when running Sysbench with a large number of threads. The function seems to return FALSE positives. rb://250	sunny	14 years ago
1993		branches/zip: Pass the file name and line number of the call... branches/zip: Pass the file name and line number of the caller of the b-tree cursor functions to the buffer pool requests, in order to make the latch diagnostics more accurate. buf_page_optimistic_get_func(): Renamed to buf_page_optimistic_get(). btr_page_get_father_node_ptr(), btr_insert_on_non_leaf_level(), btr_pcur_open(), btr_pcur_open_with_no_init(), btr_pcur_open_on_user_rec(), btr_pcur_open_at_rnd_pos(), btr_pcur_restore_position(), btr_cur_open_at_index_side(), btr_cur_open_at_rnd_pos(): Rename the function to _func and add the parameters file, line. Define wrapper macros with __FILE__, __LINE__. btr_cur_search_to_nth_level(): Add the parameters file, line.	marko	14 years ago
1992		branches/zip: buf_LRU_invalidate_tablespace(): Ensure that p... branches/zip: buf_LRU_invalidate_tablespace(): Ensure that prev_bpage is not relocated when freeing a compressed block.  This avoids the costly rescan of the LRU list.  (Bug #35077, Issue #449) At most one buffer-fix will be active at a time, affecting two blocks: the buf_page_t and the compressed page frame. This should not block the memory defragmentation in buf0buddy.c too much.  In fact, it may avoid unnecessary copying if also prev_bpage belongs to the tablespace that is being invalidated. rb://240	marko	14 years ago
1991		branches/zip: Clean up CHECK TABLE error handling. (Issue #2... branches/zip: Clean up CHECK TABLE error handling. (Issue #220) ha_innobase::change_active_index(): Clean up code formatting. ha_innobase::check(): Incorporate the code from row_check_table_for_mysql().  Report errors to the client connection instead of writing them to the error log. row_check_table_for_mysql(): Remove. row_check_index_for_mysql(): Renamed from row_scan_and_check_index(). Let the caller initialize prebuilt, and assume that the index is usable. rb://178 approved by Sunny Bains	marko	14 years ago
1990		branches/zip: Relax assertion on the number of index defined branches/zip: Relax assertion on the number of index defined in InnoDB must be comparable with that of MySQL to tolerate possible dictionary inconsistency. Fix Mantis issue #455, "UNIV_DEBUG+ assert ha_innodb.cc:3152 ib_num_index >=  mysql_num_index". rb://248 Approved by Marko.	jyang	14 years ago
1989		branches/zip: ChangeLog: Document the merge of 6471:6538 fro... branches/zip: ChangeLog: Document the merge of 6471:6538 from branches/5.1.	marko	14 years ago
1988		branches/zip: Merge revisions 6471:6538 from branches/5.1: branches/zip: Merge revisions 6471:6538 from branches/5.1:   ------------------------------------------------------------------------   r6488 | sunny | 2010-01-21 02:55:08 +0200 (Thu, 21 Jan 2010) | 2 lines   Changed paths:      M /branches/5.1/mysql-test/innodb-autoinc.result      M /branches/5.1/mysql-test/innodb-autoinc.test   branches/5.1: Factor out test for bug#44030 from innodb-autoinc.test   into a separate test/result files.   ------------------------------------------------------------------------   r6489 | sunny | 2010-01-21 02:57:50 +0200 (Thu, 21 Jan 2010) | 2 lines   Changed paths:      A /branches/5.1/mysql-test/innodb-autoinc-44030.result      A /branches/5.1/mysql-test/innodb-autoinc-44030.test   branches/5.1: Factor out test for bug#44030 from innodb-autoinc.test   into a separate test/result files.   ------------------------------------------------------------------------   r6492 | sunny | 2010-01-21 09:38:35 +0200 (Thu, 21 Jan 2010) | 1 line   Changed paths:      M /branches/5.1/mysql-test/innodb-autoinc-44030.test   branches/5.1: Add reference to bug#47621 in the comment.   ------------------------------------------------------------------------   r6535 | sunny | 2010-01-30 00:08:40 +0200 (Sat, 30 Jan 2010) | 11 lines   Changed paths:      M /branches/5.1/handler/ha_innodb.cc   branches/5.1: Undo the change from r6424. We need to return DB_SUCCESS even   if we were unable to initialize the tabe autoinc value. This is required for   the open to succeed. The only condition we currently treat as a hard error   is if the autoinc field instance passed in by MySQL is NULL.   Previously if the table autoinc value was 0 and the next value was requested   we had an assertion that would fail. Change that assertion and treat a value   of 0 to mean that the autoinc system is unavailable. Generation of next   value will now return failure.   rb://237   ------------------------------------------------------------------------   r6536 | sunny | 2010-01-30 00:13:42 +0200 (Sat, 30 Jan 2010) | 6 lines   Changed paths:      M /branches/5.1/handler/ha_innodb.cc      M /branches/5.1/mysql-test/innodb-autoinc.result      M /branches/5.1/mysql-test/innodb-autoinc.test   branches/5.1: Check *first_value everytime against the column max   value and  set *first_value to next autoinc if it's > col max value.   ie.  not rely on what is passed in from MySQL.   [49497] Error 1467 (ER_AUTOINC_READ_FAILED) on inserting a negative value   rb://236   ------------------------------------------------------------------------   r6537 | sunny | 2010-01-30 00:35:00 +0200 (Sat, 30 Jan 2010) | 2 lines   Changed paths:      M /branches/5.1/handler/ha_innodb.cc      M /branches/5.1/mysql-test/innodb-autoinc.result      M /branches/5.1/mysql-test/innodb-autoinc.test   branches/5.1: Undo r6536.   ------------------------------------------------------------------------   r6538 | sunny | 2010-01-30 00:43:06 +0200 (Sat, 30 Jan 2010) | 6 lines   Changed paths:      M /branches/5.1/handler/ha_innodb.cc      M /branches/5.1/mysql-test/innodb-autoinc.result      M /branches/5.1/mysql-test/innodb-autoinc.test   branches/5.1: Check *first_value every time against the column max   value and  set *first_value to next autoinc if it's > col max value.   ie.  not rely on what is passed in from MySQL.   [49497] Error 1467 (ER_AUTOINC_READ_FAILED) on inserting a negative value   rb://236   ------------------------------------------------------------------------	marko	14 years ago

• « Newer
• Older »

Loggerhead is a web-based interface for Breezy
Version: 2.0.1