mariadb/storage/innobase/ut/ut0rbt.cc

1329 lines
32 KiB
C++
Raw Normal View History

branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/***************************************************************************//**
Copyright (c) 2007, 2010, Oracle and/or its affiliates. All Rights Reserved.
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
This program is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free Software
Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Suite 500, Boston, MA 02110-1335 USA
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
*****************************************************************************/
/********************************************************************//**
Red-Black tree implementation
(c) 2007 Oracle/Innobase Oy
Created 2007-03-20 Sunny Bains
***********************************************************************/
#include "ut0rbt.h"
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Definition of a red-black tree
==============================
A red-black tree is a binary search tree which has the following
red-black properties:
1. Every node is either red or black.
2. Every leaf (NULL - in our case tree->nil) is black.
3. If a node is red, then both its children are black.
4. Every simple path from a node to a descendant leaf contains the
same number of black nodes.
from (3) above, the implication is that on any path from the root
to a leaf, red nodes must not be adjacent.
However, any number of black nodes may appear in a sequence.
*/
#if defined(IB_RBT_TESTING)
#warning "Testing enabled!"
#endif
#define ROOT(t) (t->root->left)
#define SIZEOF_NODE(t) ((sizeof(ib_rbt_node_t) + t->sizeof_value) - 1)
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Print out the sub-tree recursively. */
static
void
rbt_print_subtree(
/*==============*/
const ib_rbt_t* tree, /*!< in: tree to traverse */
const ib_rbt_node_t* node, /*!< in: node to print */
ib_rbt_print_node print) /*!< in: print key function */
{
/* FIXME: Doesn't do anything yet */
if (node != tree->nil) {
print(node);
rbt_print_subtree(tree, node->left, print);
rbt_print_subtree(tree, node->right, print);
}
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Verify that the keys are in order.
@return TRUE of OK. FALSE if not ordered */
static
ibool
rbt_check_ordering(
/*===============*/
const ib_rbt_t* tree) /*!< in: tree to verfify */
{
const ib_rbt_node_t* node;
const ib_rbt_node_t* prev = NULL;
/* Iterate over all the nodes, comparing each node with the prev */
for (node = rbt_first(tree); node; node = rbt_next(tree, prev)) {
if (prev) {
int result;
if (tree->cmp_arg) {
result = tree->compare_with_arg(
tree->cmp_arg, prev->value,
node->value);
} else {
result = tree->compare(
prev->value, node->value);
}
if (result >= 0) {
return(FALSE);
}
}
prev = node;
}
return(TRUE);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Check that every path from the root to the leaves has the same count.
Count is expressed in the number of black nodes.
@return 0 on failure else black height of the subtree */
static
ibool
rbt_count_black_nodes(
/*==================*/
const ib_rbt_t* tree, /*!< in: tree to verify */
const ib_rbt_node_t* node) /*!< in: start of sub-tree */
{
ulint result;
if (node != tree->nil) {
ulint left_height = rbt_count_black_nodes(tree, node->left);
ulint right_height = rbt_count_black_nodes(tree, node->right);
if (left_height == 0
|| right_height == 0
|| left_height != right_height) {
result = 0;
} else if (node->color == IB_RBT_RED) {
/* Case 3 */
if (node->left->color != IB_RBT_BLACK
|| node->right->color != IB_RBT_BLACK) {
result = 0;
} else {
result = left_height;
}
/* Check if it's anything other than RED or BLACK. */
} else if (node->color != IB_RBT_BLACK) {
result = 0;
} else {
result = right_height + 1;
}
} else {
result = 1;
}
return(result);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Turn the node's right child's left sub-tree into node's right sub-tree.
This will also make node's right child it's parent. */
static
void
rbt_rotate_left(
/*============*/
const ib_rbt_node_t* nil, /*!< in: nil node of the tree */
ib_rbt_node_t* node) /*!< in: node to rotate */
{
ib_rbt_node_t* right = node->right;
node->right = right->left;
if (right->left != nil) {
right->left->parent = node;
}
/* Right's new parent was node's parent. */
right->parent = node->parent;
/* Since root's parent is tree->nil and root->parent->left points
back to root, we can avoid the check. */
if (node == node->parent->left) {
/* Node was on the left of its parent. */
node->parent->left = right;
} else {
/* Node must have been on the right. */
node->parent->right = right;
}
/* Finally, put node on right's left. */
right->left = node;
node->parent = right;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Turn the node's left child's right sub-tree into node's left sub-tree.
This also make node's left child it's parent. */
static
void
rbt_rotate_right(
/*=============*/
const ib_rbt_node_t* nil, /*!< in: nil node of tree */
ib_rbt_node_t* node) /*!< in: node to rotate */
{
ib_rbt_node_t* left = node->left;
node->left = left->right;
if (left->right != nil) {
left->right->parent = node;
}
/* Left's new parent was node's parent. */
left->parent = node->parent;
/* Since root's parent is tree->nil and root->parent->left points
back to root, we can avoid the check. */
if (node == node->parent->right) {
/* Node was on the left of its parent. */
node->parent->right = left;
} else {
/* Node must have been on the left. */
node->parent->left = left;
}
/* Finally, put node on left's right. */
left->right = node;
node->parent = left;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Append a node to the tree. */
static
ib_rbt_node_t*
rbt_tree_add_child(
/*===============*/
const ib_rbt_t* tree,
ib_rbt_bound_t* parent,
ib_rbt_node_t* node)
{
/* Cast away the const. */
ib_rbt_node_t* last = (ib_rbt_node_t*) parent->last;
if (last == tree->root || parent->result < 0) {
last->left = node;
} else {
/* FIXME: We don't handle duplicates (yet)! */
ut_a(parent->result != 0);
last->right = node;
}
node->parent = last;
return(node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Generic binary tree insert */
static
ib_rbt_node_t*
rbt_tree_insert(
/*============*/
ib_rbt_t* tree,
const void* key,
ib_rbt_node_t* node)
{
ib_rbt_bound_t parent;
ib_rbt_node_t* current = ROOT(tree);
parent.result = 0;
parent.last = tree->root;
/* Regular binary search. */
while (current != tree->nil) {
parent.last = current;
if (tree->cmp_arg) {
parent.result = tree->compare_with_arg(
tree->cmp_arg, key, current->value);
} else {
parent.result = tree->compare(key, current->value);
}
if (parent.result < 0) {
current = current->left;
} else {
current = current->right;
}
}
ut_a(current == tree->nil);
rbt_tree_add_child(tree, &parent, node);
return(node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Balance a tree after inserting a node. */
static
void
rbt_balance_tree(
/*=============*/
const ib_rbt_t* tree, /*!< in: tree to balance */
ib_rbt_node_t* node) /*!< in: node that was inserted */
{
const ib_rbt_node_t* nil = tree->nil;
ib_rbt_node_t* parent = node->parent;
/* Restore the red-black property. */
node->color = IB_RBT_RED;
while (node != ROOT(tree) && parent->color == IB_RBT_RED) {
ib_rbt_node_t* grand_parent = parent->parent;
if (parent == grand_parent->left) {
ib_rbt_node_t* uncle = grand_parent->right;
if (uncle->color == IB_RBT_RED) {
/* Case 1 - change the colors. */
uncle->color = IB_RBT_BLACK;
parent->color = IB_RBT_BLACK;
grand_parent->color = IB_RBT_RED;
/* Move node up the tree. */
node = grand_parent;
} else {
if (node == parent->right) {
/* Right is a black node and node is
to the right, case 2 - move node
up and rotate. */
node = parent;
rbt_rotate_left(nil, node);
}
grand_parent = node->parent->parent;
/* Case 3. */
node->parent->color = IB_RBT_BLACK;
grand_parent->color = IB_RBT_RED;
rbt_rotate_right(nil, grand_parent);
}
} else {
ib_rbt_node_t* uncle = grand_parent->left;
if (uncle->color == IB_RBT_RED) {
/* Case 1 - change the colors. */
uncle->color = IB_RBT_BLACK;
parent->color = IB_RBT_BLACK;
grand_parent->color = IB_RBT_RED;
/* Move node up the tree. */
node = grand_parent;
} else {
if (node == parent->left) {
/* Left is a black node and node is to
the right, case 2 - move node up and
rotate. */
node = parent;
rbt_rotate_right(nil, node);
}
grand_parent = node->parent->parent;
/* Case 3. */
node->parent->color = IB_RBT_BLACK;
grand_parent->color = IB_RBT_RED;
rbt_rotate_left(nil, grand_parent);
}
}
parent = node->parent;
}
/* Color the root black. */
ROOT(tree)->color = IB_RBT_BLACK;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find the given node's successor.
@return successor node or NULL if no successor */
static
ib_rbt_node_t*
rbt_find_successor(
/*===============*/
const ib_rbt_t* tree, /*!< in: rb tree */
const ib_rbt_node_t* current) /*!< in: this is declared const
because it can be called via
rbt_next() */
{
const ib_rbt_node_t* nil = tree->nil;
ib_rbt_node_t* next = current->right;
/* Is there a sub-tree to the right that we can follow. */
if (next != nil) {
/* Follow the left most links of the current right child. */
while (next->left != nil) {
next = next->left;
}
} else { /* We will have to go up the tree to find the successor. */
ib_rbt_node_t* parent = current->parent;
/* Cast away the const. */
next = (ib_rbt_node_t*) current;
while (parent != tree->root && next == parent->right) {
next = parent;
parent = next->parent;
}
next = (parent == tree->root) ? NULL : parent;
}
return(next);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find the given node's precedecessor.
@return predecessor node or NULL if no predecesor */
static
ib_rbt_node_t*
rbt_find_predecessor(
/*=================*/
const ib_rbt_t* tree, /*!< in: rb tree */
const ib_rbt_node_t* current) /*!< in: this is declared const
because it can be called via
rbt_prev() */
{
const ib_rbt_node_t* nil = tree->nil;
ib_rbt_node_t* prev = current->left;
/* Is there a sub-tree to the left that we can follow. */
if (prev != nil) {
/* Follow the right most links of the current left child. */
while (prev->right != nil) {
prev = prev->right;
}
} else { /* We will have to go up the tree to find the precedecessor. */
ib_rbt_node_t* parent = current->parent;
/* Cast away the const. */
prev = (ib_rbt_node_t*) current;
while (parent != tree->root && prev == parent->left) {
prev = parent;
parent = prev->parent;
}
prev = (parent == tree->root) ? NULL : parent;
}
return(prev);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Replace node with child. After applying transformations eject becomes
an orphan. */
static
void
rbt_eject_node(
/*===========*/
ib_rbt_node_t* eject, /*!< in: node to eject */
ib_rbt_node_t* node) /*!< in: node to replace with */
{
/* Update the to be ejected node's parent's child pointers. */
if (eject->parent->left == eject) {
eject->parent->left = node;
} else if (eject->parent->right == eject) {
eject->parent->right = node;
} else {
ut_a(0);
}
/* eject is now an orphan but otherwise its pointers
and color are left intact. */
node->parent = eject->parent;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Replace a node with another node. */
static
void
rbt_replace_node(
/*=============*/
ib_rbt_node_t* replace, /*!< in: node to replace */
ib_rbt_node_t* node) /*!< in: node to replace with */
{
ib_rbt_color_t color = node->color;
/* Update the node pointers. */
node->left = replace->left;
node->right = replace->right;
/* Update the child node pointers. */
node->left->parent = node;
node->right->parent = node;
/* Make the parent of replace point to node. */
rbt_eject_node(replace, node);
/* Swap the colors. */
node->color = replace->color;
replace->color = color;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Detach node from the tree replacing it with one of it's children.
@return the child node that now occupies the position of the detached node */
static
ib_rbt_node_t*
rbt_detach_node(
/*============*/
const ib_rbt_t* tree, /*!< in: rb tree */
ib_rbt_node_t* node) /*!< in: node to detach */
{
ib_rbt_node_t* child;
const ib_rbt_node_t* nil = tree->nil;
if (node->left != nil && node->right != nil) {
/* Case where the node to be deleted has two children. */
ib_rbt_node_t* successor = rbt_find_successor(tree, node);
ut_a(successor != nil);
ut_a(successor->parent != nil);
ut_a(successor->left == nil);
child = successor->right;
/* Remove the successor node and replace with its child. */
rbt_eject_node(successor, child);
/* Replace the node to delete with its successor node. */
rbt_replace_node(node, successor);
} else {
ut_a(node->left == nil || node->right == nil);
child = (node->left != nil) ? node->left : node->right;
/* Replace the node to delete with one of it's children. */
rbt_eject_node(node, child);
}
/* Reset the node links. */
node->parent = node->right = node->left = tree->nil;
return(child);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Rebalance the right sub-tree after deletion.
@return node to rebalance if more rebalancing required else NULL */
static
ib_rbt_node_t*
rbt_balance_right(
/*==============*/
const ib_rbt_node_t* nil, /*!< in: rb tree nil node */
ib_rbt_node_t* parent, /*!< in: parent node */
ib_rbt_node_t* sibling) /*!< in: sibling node */
{
ib_rbt_node_t* node = NULL;
ut_a(sibling != nil);
/* Case 3. */
if (sibling->color == IB_RBT_RED) {
parent->color = IB_RBT_RED;
sibling->color = IB_RBT_BLACK;
rbt_rotate_left(nil, parent);
sibling = parent->right;
ut_a(sibling != nil);
}
/* Since this will violate case 3 because of the change above. */
if (sibling->left->color == IB_RBT_BLACK
&& sibling->right->color == IB_RBT_BLACK) {
node = parent; /* Parent needs to be rebalanced too. */
sibling->color = IB_RBT_RED;
} else {
if (sibling->right->color == IB_RBT_BLACK) {
ut_a(sibling->left->color == IB_RBT_RED);
sibling->color = IB_RBT_RED;
sibling->left->color = IB_RBT_BLACK;
rbt_rotate_right(nil, sibling);
sibling = parent->right;
ut_a(sibling != nil);
}
sibling->color = parent->color;
sibling->right->color = IB_RBT_BLACK;
parent->color = IB_RBT_BLACK;
rbt_rotate_left(nil, parent);
}
return(node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Rebalance the left sub-tree after deletion.
@return node to rebalance if more rebalancing required else NULL */
static
ib_rbt_node_t*
rbt_balance_left(
/*=============*/
const ib_rbt_node_t* nil, /*!< in: rb tree nil node */
ib_rbt_node_t* parent, /*!< in: parent node */
ib_rbt_node_t* sibling) /*!< in: sibling node */
{
ib_rbt_node_t* node = NULL;
ut_a(sibling != nil);
/* Case 3. */
if (sibling->color == IB_RBT_RED) {
parent->color = IB_RBT_RED;
sibling->color = IB_RBT_BLACK;
rbt_rotate_right(nil, parent);
sibling = parent->left;
ut_a(sibling != nil);
}
/* Since this will violate case 3 because of the change above. */
if (sibling->right->color == IB_RBT_BLACK
&& sibling->left->color == IB_RBT_BLACK) {
node = parent; /* Parent needs to be rebalanced too. */
sibling->color = IB_RBT_RED;
} else {
if (sibling->left->color == IB_RBT_BLACK) {
ut_a(sibling->right->color == IB_RBT_RED);
sibling->color = IB_RBT_RED;
sibling->right->color = IB_RBT_BLACK;
rbt_rotate_left(nil, sibling);
sibling = parent->left;
ut_a(sibling != nil);
}
sibling->color = parent->color;
sibling->left->color = IB_RBT_BLACK;
parent->color = IB_RBT_BLACK;
rbt_rotate_right(nil, parent);
}
return(node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Delete the node and rebalance the tree if necessary */
static
void
rbt_remove_node_and_rebalance(
/*==========================*/
ib_rbt_t* tree, /*!< in: rb tree */
ib_rbt_node_t* node) /*!< in: node to remove */
{
/* Detach node and get the node that will be used
as rebalance start. */
ib_rbt_node_t* child = rbt_detach_node(tree, node);
if (node->color == IB_RBT_BLACK) {
ib_rbt_node_t* last = child;
ROOT(tree)->color = IB_RBT_RED;
while (child && child->color == IB_RBT_BLACK) {
ib_rbt_node_t* parent = child->parent;
/* Did the deletion cause an imbalance in the
parents left sub-tree. */
if (parent->left == child) {
child = rbt_balance_right(
tree->nil, parent, parent->right);
} else if (parent->right == child) {
child = rbt_balance_left(
tree->nil, parent, parent->left);
} else {
ut_error;
}
if (child) {
last = child;
}
}
ut_a(last);
last->color = IB_RBT_BLACK;
ROOT(tree)->color = IB_RBT_BLACK;
}
/* Note that we have removed a node from the tree. */
--tree->n_nodes;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Recursively free the nodes. */
static
void
rbt_free_node(
/*==========*/
ib_rbt_node_t* node, /*!< in: node to free */
ib_rbt_node_t* nil) /*!< in: rb tree nil node */
{
if (node != nil) {
rbt_free_node(node->left, nil);
rbt_free_node(node->right, nil);
ut_free(node);
}
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Free all the nodes and free the tree. */
UNIV_INTERN
void
rbt_free(
/*=====*/
ib_rbt_t* tree) /*!< in: rb tree to free */
{
rbt_free_node(tree->root, tree->nil);
ut_free(tree->nil);
ut_free(tree);
}
/**********************************************************************//**
Create an instance of a red black tree, whose comparison function takes
an argument
@return an empty rb tree */
UNIV_INTERN
ib_rbt_t*
rbt_create_arg_cmp(
/*===============*/
size_t sizeof_value, /*!< in: sizeof data item */
ib_rbt_arg_compare
compare, /*!< in: fn to compare items */
const void* cmp_arg) /*!< in: compare fn arg */
{
ib_rbt_t* tree;
ut_a(cmp_arg);
tree = rbt_create(sizeof_value, NULL);
tree->cmp_arg = cmp_arg;
tree->compare_with_arg = compare;
return(tree);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Create an instance of a red black tree.
@return an empty rb tree */
UNIV_INTERN
ib_rbt_t*
rbt_create(
/*=======*/
size_t sizeof_value, /*!< in: sizeof data item */
ib_rbt_compare compare) /*!< in: fn to compare items */
{
ib_rbt_t* tree;
ib_rbt_node_t* node;
tree = (ib_rbt_t*) ut_malloc(sizeof(*tree));
memset(tree, 0, sizeof(*tree));
tree->sizeof_value = sizeof_value;
/* Create the sentinel (NIL) node. */
node = tree->nil = (ib_rbt_node_t*) ut_malloc(sizeof(*node));
memset(node, 0, sizeof(*node));
node->color = IB_RBT_BLACK;
node->parent = node->left = node->right = node;
/* Create the "fake" root, the real root node will be the
left child of this node. */
node = tree->root = (ib_rbt_node_t*) ut_malloc(sizeof(*node));
memset(node, 0, sizeof(*node));
node->color = IB_RBT_BLACK;
node->parent = node->left = node->right = tree->nil;
tree->compare = compare;
return(tree);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Generic insert of a value in the rb tree.
@return inserted node */
UNIV_INTERN
const ib_rbt_node_t*
rbt_insert(
/*=======*/
ib_rbt_t* tree, /*!< in: rb tree */
const void* key, /*!< in: key for ordering */
const void* value) /*!< in: value of key, this value
is copied to the node */
{
ib_rbt_node_t* node;
/* Create the node that will hold the value data. */
node = (ib_rbt_node_t*) ut_malloc(SIZEOF_NODE(tree));
memcpy(node->value, value, tree->sizeof_value);
node->parent = node->left = node->right = tree->nil;
/* Insert in the tree in the usual way. */
rbt_tree_insert(tree, key, node);
rbt_balance_tree(tree, node);
++tree->n_nodes;
return(node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Add a new node to the tree, useful for data that is pre-sorted.
@return appended node */
UNIV_INTERN
const ib_rbt_node_t*
rbt_add_node(
/*=========*/
ib_rbt_t* tree, /*!< in: rb tree */
ib_rbt_bound_t* parent, /*!< in: bounds */
const void* value) /*!< in: this value is copied
to the node */
{
ib_rbt_node_t* node;
/* Create the node that will hold the value data */
node = (ib_rbt_node_t*) ut_malloc(SIZEOF_NODE(tree));
memcpy(node->value, value, tree->sizeof_value);
node->parent = node->left = node->right = tree->nil;
/* If tree is empty */
if (parent->last == NULL) {
parent->last = tree->root;
}
/* Append the node, the hope here is that the caller knows
what s/he is doing. */
rbt_tree_add_child(tree, parent, node);
rbt_balance_tree(tree, node);
++tree->n_nodes;
#if defined(IB_RBT_TESTING)
ut_a(rbt_validate(tree));
#endif
return(node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find a matching node in the rb tree.
@return NULL if not found else the node where key was found */
UNIV_INTERN
const ib_rbt_node_t*
rbt_lookup(
/*=======*/
const ib_rbt_t* tree, /*!< in: rb tree */
const void* key) /*!< in: key to use for search */
{
const ib_rbt_node_t* current = ROOT(tree);
/* Regular binary search. */
while (current != tree->nil) {
int result;
if (tree->cmp_arg) {
result = tree->compare_with_arg(
tree->cmp_arg, key, current->value);
} else {
result = tree->compare(key, current->value);
}
if (result < 0) {
current = current->left;
} else if (result > 0) {
current = current->right;
} else {
break;
}
}
return(current != tree->nil ? current : NULL);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Delete a node indentified by key.
@return TRUE if success FALSE if not found */
UNIV_INTERN
ibool
rbt_delete(
/*=======*/
ib_rbt_t* tree, /*!< in: rb tree */
const void* key) /*!< in: key to delete */
{
ibool deleted = FALSE;
ib_rbt_node_t* node = (ib_rbt_node_t*) rbt_lookup(tree, key);
if (node) {
rbt_remove_node_and_rebalance(tree, node);
ut_free(node);
deleted = TRUE;
}
return(deleted);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Remove a node from the rb tree, the node is not free'd, that is the
callers responsibility.
@return deleted node but without the const */
UNIV_INTERN
ib_rbt_node_t*
rbt_remove_node(
/*============*/
ib_rbt_t* tree, /*!< in: rb tree */
const ib_rbt_node_t* const_node) /*!< in: node to delete, this
is a fudge and declared const
because the caller can access
only const nodes */
{
/* Cast away the const. */
rbt_remove_node_and_rebalance(tree, (ib_rbt_node_t*) const_node);
/* This is to make it easier to do something like this:
ut_free(rbt_remove_node(node));
*/
return((ib_rbt_node_t*) const_node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find the node that has the lowest key that is >= key.
@return node satisfying the lower bound constraint or NULL */
UNIV_INTERN
const ib_rbt_node_t*
rbt_lower_bound(
/*============*/
const ib_rbt_t* tree, /*!< in: rb tree */
const void* key) /*!< in: key to search */
{
ib_rbt_node_t* lb_node = NULL;
ib_rbt_node_t* current = ROOT(tree);
while (current != tree->nil) {
int result;
if (tree->cmp_arg) {
result = tree->compare_with_arg(
tree->cmp_arg, key, current->value);
} else {
result = tree->compare(key, current->value);
}
if (result > 0) {
current = current->right;
} else if (result < 0) {
lb_node = current;
current = current->left;
} else {
lb_node = current;
break;
}
}
return(lb_node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find the node that has the greatest key that is <= key.
@return node satisfying the upper bound constraint or NULL */
UNIV_INTERN
const ib_rbt_node_t*
rbt_upper_bound(
/*============*/
const ib_rbt_t* tree, /*!< in: rb tree */
const void* key) /*!< in: key to search */
{
ib_rbt_node_t* ub_node = NULL;
ib_rbt_node_t* current = ROOT(tree);
while (current != tree->nil) {
int result;
if (tree->cmp_arg) {
result = tree->compare_with_arg(
tree->cmp_arg, key, current->value);
} else {
result = tree->compare(key, current->value);
}
if (result > 0) {
ub_node = current;
current = current->right;
} else if (result < 0) {
current = current->left;
} else {
ub_node = current;
break;
}
}
return(ub_node);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find the node that has the greatest key that is <= key.
@return value of result */
UNIV_INTERN
int
rbt_search(
/*=======*/
const ib_rbt_t* tree, /*!< in: rb tree */
ib_rbt_bound_t* parent, /*!< in: search bounds */
const void* key) /*!< in: key to search */
{
ib_rbt_node_t* current = ROOT(tree);
/* Every thing is greater than the NULL root. */
parent->result = 1;
parent->last = NULL;
while (current != tree->nil) {
parent->last = current;
if (tree->cmp_arg) {
parent->result = tree->compare_with_arg(
tree->cmp_arg, key, current->value);
} else {
parent->result = tree->compare(key, current->value);
}
if (parent->result > 0) {
current = current->right;
} else if (parent->result < 0) {
current = current->left;
} else {
break;
}
}
return(parent->result);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Find the node that has the greatest key that is <= key. But use the
supplied comparison function.
@return value of result */
UNIV_INTERN
int
rbt_search_cmp(
/*===========*/
const ib_rbt_t* tree, /*!< in: rb tree */
ib_rbt_bound_t* parent, /*!< in: search bounds */
const void* key, /*!< in: key to search */
ib_rbt_compare compare, /*!< in: fn to compare items */
ib_rbt_arg_compare
arg_compare) /*!< in: fn to compare items
with argument */
{
ib_rbt_node_t* current = ROOT(tree);
/* Every thing is greater than the NULL root. */
parent->result = 1;
parent->last = NULL;
while (current != tree->nil) {
parent->last = current;
if (arg_compare) {
ut_ad(tree->cmp_arg);
parent->result = arg_compare(
tree->cmp_arg, key, current->value);
} else {
parent->result = compare(key, current->value);
}
if (parent->result > 0) {
current = current->right;
} else if (parent->result < 0) {
current = current->left;
} else {
break;
}
}
return(parent->result);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Return the left most node in the tree. */
UNIV_INTERN
const ib_rbt_node_t*
rbt_first(
/*======*/
/* out leftmost node or NULL */
const ib_rbt_t* tree) /* in: rb tree */
{
ib_rbt_node_t* first = NULL;
ib_rbt_node_t* current = ROOT(tree);
while (current != tree->nil) {
first = current;
current = current->left;
}
return(first);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Return the right most node in the tree.
@return the rightmost node or NULL */
UNIV_INTERN
const ib_rbt_node_t*
rbt_last(
/*=====*/
const ib_rbt_t* tree) /*!< in: rb tree */
{
ib_rbt_node_t* last = NULL;
ib_rbt_node_t* current = ROOT(tree);
while (current != tree->nil) {
last = current;
current = current->right;
}
return(last);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Return the next node.
@return node next from current */
UNIV_INTERN
const ib_rbt_node_t*
rbt_next(
/*=====*/
const ib_rbt_t* tree, /*!< in: rb tree */
const ib_rbt_node_t* current) /*!< in: current node */
{
return(current ? rbt_find_successor(tree, current) : NULL);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Return the previous node.
@return node prev from current */
UNIV_INTERN
const ib_rbt_node_t*
rbt_prev(
/*=====*/
const ib_rbt_t* tree, /*!< in: rb tree */
const ib_rbt_node_t* current) /*!< in: current node */
{
return(current ? rbt_find_predecessor(tree, current) : NULL);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Reset the tree. Delete all the nodes. */
UNIV_INTERN
void
rbt_clear(
/*======*/
ib_rbt_t* tree) /*!< in: rb tree */
{
rbt_free_node(ROOT(tree), tree->nil);
tree->n_nodes = 0;
tree->root->left = tree->root->right = tree->nil;
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Merge the node from dst into src. Return the number of nodes merged.
@return no. of recs merged */
UNIV_INTERN
ulint
rbt_merge_uniq(
/*===========*/
ib_rbt_t* dst, /*!< in: dst rb tree */
const ib_rbt_t* src) /*!< in: src rb tree */
{
ib_rbt_bound_t parent;
ulint n_merged = 0;
const ib_rbt_node_t* src_node = rbt_first(src);
if (rbt_empty(src) || dst == src) {
return(0);
}
for (/* No op */; src_node; src_node = rbt_next(src, src_node)) {
if (rbt_search(dst, &parent, src_node->value) != 0) {
rbt_add_node(dst, &parent, src_node->value);
++n_merged;
}
}
return(n_merged);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Merge the node from dst into src. Return the number of nodes merged.
Delete the nodes from src after copying node to dst. As a side effect
the duplicates will be left untouched in the src.
@return no. of recs merged */
UNIV_INTERN
ulint
rbt_merge_uniq_destructive(
/*=======================*/
ib_rbt_t* dst, /*!< in: dst rb tree */
ib_rbt_t* src) /*!< in: src rb tree */
{
ib_rbt_bound_t parent;
ib_rbt_node_t* src_node;
ulint old_size = rbt_size(dst);
if (rbt_empty(src) || dst == src) {
return(0);
}
for (src_node = (ib_rbt_node_t*) rbt_first(src); src_node; /* */) {
ib_rbt_node_t* prev = src_node;
src_node = (ib_rbt_node_t*) rbt_next(src, prev);
/* Skip duplicates. */
if (rbt_search(dst, &parent, prev->value) != 0) {
/* Remove and reset the node but preserve
the node (data) value. */
rbt_remove_node_and_rebalance(src, prev);
/* The nil should be taken from the dst tree. */
prev->parent = prev->left = prev->right = dst->nil;
rbt_tree_add_child(dst, &parent, prev);
rbt_balance_tree(dst, prev);
++dst->n_nodes;
}
}
#if defined(IB_RBT_TESTING)
ut_a(rbt_validate(dst));
ut_a(rbt_validate(src));
#endif
return(rbt_size(dst) - old_size);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Check that every path from the root to the leaves has the same count and
the tree nodes are in order.
@return TRUE if OK FALSE otherwise */
UNIV_INTERN
ibool
rbt_validate(
/*=========*/
const ib_rbt_t* tree) /*!< in: RB tree to validate */
{
if (rbt_count_black_nodes(tree, ROOT(tree)) > 0) {
return(rbt_check_ordering(tree));
}
return(FALSE);
}
branches/innodb+ Merge r6915:6992 from branches/innodb+multipbp (i.e.: all the changes made since it's creation) This also reverts r6930 to branches/innodb+ because a different solution for that issue is already present in innodb+multibp which is being merged. After this commit branches/innodb+multibp should be discarded and this branch should become our main development tree. ------------------------------------------------------------------------ r6915 | sbains | 2010-03-31 07:33:43 +0300 (Wed, 31 Mar 2010) | 1 line Changed paths: A /branches/innodb+multibp (from /branches/innodb+:6914) Creating a branch for the multiple buffer pool ------------------------------------------------------------------------ r6916 | sbains | 2010-03-31 08:21:00 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/btr/btr0btr.c M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/btr/btr0sea.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/buf/buf0rea.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/handler/i_s.cc M /branches/innodb+multibp/ibuf/ibuf0ibuf.c M /branches/innodb+multibp/include/buf0buddy.h M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/buf0lru.h M /branches/innodb+multibp/include/buf0rea.h M /branches/innodb+multibp/include/buf0types.h M /branches/innodb+multibp/include/ibuf0ibuf.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/univ.i M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mem/mem0mem.c M /branches/innodb+multibp/page/page0zip.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/trx/trx0trx.c M /branches/innodb+multibp/trx/trx0undo.c branches/innodb+multibp: Unable to crash it with UNIV_DEBUG and UNIV_SYNC_DEBUG with both ibtests and Sysbench. The patch now needs a workout from Michael. ------------------------------------------------------------------------ r6917 | sbains | 2010-03-31 08:56:18 +0300 (Wed, 31 Mar 2010) | 2 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix error introduced in r6916. ------------------------------------------------------------------------ r6923 | sbains | 2010-03-31 15:16:04 +0300 (Wed, 31 Mar 2010) | 3 lines Changed paths: M /branches/innodb+multibp/btr/btr0cur.c M /branches/innodb+multibp/buf/buf0buddy.c M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/buf/buf0lru.c M /branches/innodb+multibp/include/buf0buddy.ic M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/page/page0zip.c branches/innodb+multibp: Fix whitespace issues. Add function buf_pool_from_block(). Add some comments to parameters. ------------------------------------------------------------------------ r6932 | sbains | 2010-04-01 01:12:07 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/univ.i branches/innodb+multibp: Remove bogus assertion. It's possible for the space and offset of a page to be undefined during the lifecycle of a page. Remove the debug #defines from univ.i. ------------------------------------------------------------------------ r6933 | sbains | 2010-04-01 01:22:40 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/srv/srv0start.c branches/innodb+multibp: Fix whitespace issues. ------------------------------------------------------------------------ r6934 | sbains | 2010-04-01 01:53:18 +0300 (Thu, 01 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/CMakeLists.txt M /branches/innodb+multibp/ChangeLog M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/handler/ha_innodb.cc M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/srv0srv.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/include/trx0purge.h M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/mysql-test/innodb_bug38231.test A /branches/innodb+multibp/mysql-test/innodb_bug51920.result (from /branches/innodb+/mysql-test/innodb_bug51920.result:6931) A /branches/innodb+multibp/mysql-test/innodb_bug51920.test (from /branches/innodb+/mysql-test/innodb_bug51920.test:6931) M /branches/innodb+multibp/row/row0sel.c M /branches/innodb+multibp/srv/srv0srv.c M /branches/innodb+multibp/srv/srv0start.c M /branches/innodb+multibp/sync/sync0sync.c M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: Merge revisions r6914:6931 from branches/innodb+ ------------------------------------------------------------------------ r6935 | sbains | 2010-04-01 02:08:32 +0300 (Thu, 01 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/mtr/mtr0mtr.c branches/innodb+multibp: Fix the debug assertions for flush order mutex. These were missed in r6934. ------------------------------------------------------------------------ r6936 | sbains | 2010-04-01 02:46:52 +0300 (Thu, 01 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp: Because now we have multiple instances of a mutex at the same level and these mutexes can be acquired simultaneously we can't simply check for <= level. We need to check for <= level - 1. ------------------------------------------------------------------------ r6937 | sbains | 2010-04-01 04:40:17 +0300 (Thu, 01 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/trx/trx0purge.c branches/innodb+multibp: We need to check if the history list len is > than some threshold not that it is evenly divisible by the some batch size. While running tests on dscczz01 I've observed that the purge thread can't keep up with the generation of the UNDO log records because of the faster code. ------------------------------------------------------------------------ r6938 | irana | 2010-04-01 10:15:00 +0300 (Thu, 01 Apr 2010) | 7 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c M /branches/innodb+multibp/buf/buf0flu.c M /branches/innodb+multibp/include/buf0buf.h M /branches/innodb+multibp/include/buf0buf.ic M /branches/innodb+multibp/include/buf0flu.h M /branches/innodb+multibp/include/buf0flu.ic M /branches/innodb+multibp/include/log0log.h M /branches/innodb+multibp/include/sync0sync.h M /branches/innodb+multibp/log/log0log.c M /branches/innodb+multibp/log/log0recv.c M /branches/innodb+multibp/mtr/mtr0mtr.c M /branches/innodb+multibp/sync/sync0sync.c branches/innodb+multibp The buf_flush_order patch that was ported in from 1.1 won't work with multiple buffer pools. This patch moves the mutex protecting order of insertion in the flush list(s) to log_sys struct so that we can have one global mutex protecting insertions into all flush list(s) ------------------------------------------------------------------------ r6941 | sbains | 2010-04-02 00:51:28 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp: We should get the record heap no to check recursively only if we are checking a record lock. Prior to this fix we were doing it for table locks as well, this is a bug. ------------------------------------------------------------------------ r6942 | csun | 2010-04-02 02:39:10 +0300 (Fri, 02 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/ha/ha0ha.c branches/innodb+multibp: fix compiler errors on Windows. Move ut_ad() to after declarations for C file. ------------------------------------------------------------------------ r6943 | sbains | 2010-04-03 05:14:25 +0300 (Sat, 03 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: Remove the code that created the fake buffer pool. ------------------------------------------------------------------------ r6945 | irana | 2010-04-05 23:35:29 +0300 (Mon, 05 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Revert r6941 as it does not resolve the issue and we have to take back the whole fix for bug#49047 ------------------------------------------------------------------------ r6946 | irana | 2010-04-05 23:50:42 +0300 (Mon, 05 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/include/ut0ut.h M /branches/innodb+multibp/include/ut0ut.ic M /branches/innodb+multibp/lock/lock0lock.c branches/innodb+multibp Merged revisions 6932:6944 from branches/innodb+ This solely includes the reversal of fix for bug#49047 ------------------------------------------------------------------------ r6947 | sbains | 2010-04-06 01:33:46 +0300 (Tue, 06 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/buf/buf0lru.c branches/innodb+multibp: Remove the log sys mutex acquisition when doing buffer pool stat aggregation. A dirty read here should suffice. ------------------------------------------------------------------------ r6951 | irana | 2010-04-06 17:25:29 +0300 (Tue, 06 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+mbp Initialize the buf_page_t::buf_pool pointer when the descriptor is allocated using buf_buddy_alloc(). ------------------------------------------------------------------------ r6954 | jyang | 2010-04-06 21:24:46 +0300 (Tue, 06 Apr 2010) | 4 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix a possible null pointer of index_mapping in a race condition. ------------------------------------------------------------------------ r6958 | sbains | 2010-04-07 00:27:44 +0300 (Wed, 07 Apr 2010) | 3 lines Changed paths: M /branches/innodb+multibp/include/ut0mem.h M /branches/innodb+multibp/ut/ut0mem.c branches/innodb+multibp: Fix part of Bug#52546. We allow ut_free() to accept a NULL pointer and treat it as a nop. ------------------------------------------------------------------------ r6961 | jyang | 2010-04-07 10:50:03 +0300 (Wed, 07 Apr 2010) | 9 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp: Fix for bug #52580: Crash in ha_innobase::open on executing INSERT with concurrent ALTER TABLE. Change in MySQL bug #51557 releases the mutex LOCK_open before ha_innobase::open(), causing racing condition for index translation table creation. Fix it by adding dict_sys mutex for the operation. rb://283, approved by Marko. ------------------------------------------------------------------------ r6963 | irana | 2010-04-07 19:14:10 +0300 (Wed, 07 Apr 2010) | 15 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Force setting of buf_pool->LRU_old_ratio by calling buf_LRU_old_ratio_update() with adjust set to TRUE. This will make sure that we grab the buf_pool mutex and actually adjust the buf_pool->LRU_old pointer instead of just updating the buf_pool->LRU_old_ratio. Note that after this change there is no call to buf_LRU_old_ratio_update() with adjust set to FALSE and therefore this parameter should be removed. I am keeping it for now to first make sure that the fix does work. Approved by: No one. Sunny agreed with my hypothesis of the problem. ------------------------------------------------------------------------ r6964 | irana | 2010-04-07 19:59:59 +0300 (Wed, 07 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/handler/ha_innodb.cc branches/innodb+multibp Remove a too strong assertion on behalf of Jimmy. ------------------------------------------------------------------------ r6971 | sbains | 2010-04-09 13:23:33 +0300 (Fri, 09 Apr 2010) | 6 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When getting the oldest (minimum) LSN value from all the flush lists we need to acquire the flush list mutex. We were incorrectly acquiring the buffer pool mutex. This patch should fix a slew of bugs reported by Michael. ------------------------------------------------------------------------ r6972 | sbains | 2010-04-10 00:25:09 +0300 (Sat, 10 Apr 2010) | 5 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: We should not reset the lsn to 0 when we encounter an empty flush list. Oldest LSN should be 0 only when all flush lists are empty. e.g., without this fix if even one flush list was empty we would end up breaking WAL. ------------------------------------------------------------------------ r6987 | sbains | 2010-04-14 00:14:13 +0300 (Wed, 14 Apr 2010) | 12 lines Changed paths: M /branches/innodb+multibp/buf/buf0buf.c branches/innodb+multibp: When calculating the oldest_lsn we can have a situation where we've iterated to say buffer pool 3 and another thread adds two new dirty pages, the first to buffer pool 1 and the second to buffer pool 4. Up to say buffer pool 3 the oldest_lsn was 0. Now, we will end up returning the lsn at buffer pool 4 as the oldest LSN. We prevent this by acquiring the flush order mutex. One other future option is to calculate the min_lsn when flushing pages from the list and maintaining a running total using atomics. That way we can get rid of this function altogether. The atomics will only really be required when we do parallel flushing. ------------------------------------------------------------------------ r6992 | sbains | 2010-04-14 02:45:59 +0300 (Wed, 14 Apr 2010) | 2 lines Changed paths: M /branches/innodb+multibp/include/ut0rbt.h M /branches/innodb+multibp/ut/ut0rbt.c branches/innodb+multibp: Fix copyright of the rbt code. ------------------------------------------------------------------------
2010-04-19 15:44:15 +00:00
/**********************************************************************//**
Iterate over the tree in depth first order. */
UNIV_INTERN
void
rbt_print(
/*======*/
const ib_rbt_t* tree, /*!< in: tree to traverse */
ib_rbt_print_node print) /*!< in: print function */
{
rbt_print_subtree(tree, ROOT(tree), print);
}