From: Dmitry Lenev Date: October 24 2011 7:42am Subject: bzr push into mysql-5.5 branch (Dmitry.Lenev:3588 to 3589) Bug#13116518 List-Archive: http://lists.mysql.com/commits/141564 X-Bug: 13116518 Message-Id: <20111024074238.30E6274362E@bandersnatch> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit 3589 Dmitry Lenev 2011-10-23 Fix for bug #13116518 - "OPEN_TABLES() SHOULD NOT ALLOCATE AND FREE NEW_FRM_MEM WITHOUT NEEDING TO". During the process of opening tables for a statement, we allocated memory which was used only during view loading even in cases when the statement didn't use any views. Such an unnecessary allocation (and corresponding freeing) might have caused significant performance overhead in some workloads. For example, it caused up to 15% slowdown in a simple stored routine calculating Fibonacci's numbers. This memory was pre-allocated as part of "new_frm_mem" MEM_ROOT initialization at the beginning of open_tables(). This patch addresses this issue by turning off memory pre-allocation during initialization for this MEM_ROOT. Now, memory on this root will be allocated only at the point when the first .FRM for a view is opened. The patch doesn't contain a test case since it is hard to test the performance improvements or the absence of memory allocation in our test framework. modified: sql/sql_base.cc 3588 Ashish Agarwal 2011-10-22 [merge] bug#11758979 - 51252: ARCHIVE TABLES STILL FAIL UNDER STRESS TESTS: CRASH, CORRUPTION, 4G MEMOR Issue: Valgrind errors due to checksum and optimize query against archive tables with null columns. Table record buffer was not initialized. Solution: Initialize the record buffer. modified: mysql-test/r/archive.result mysql-test/t/archive.test storage/archive/ha_archive.cc === modified file 'sql/sql_base.cc' --- a/sql/sql_base.cc 2011-08-24 12:22:11 +0000 +++ b/sql/sql_base.cc 2011-10-23 05:37:35 +0000 @@ -4810,10 +4810,11 @@ bool open_tables(THD *thd, TABLE_LIST ** } /* - temporary mem_root for new .frm parsing. - TODO: variables for size + Initialize temporary MEM_ROOT for new .FRM parsing. Do not allocate + anything yet, to avoid penalty for statements which don't use views + and thus new .FRM format. */ - init_sql_alloc(&new_frm_mem, 8024, 8024); + init_sql_alloc(&new_frm_mem, 8024, 0); thd->current_tablenr= 0; restart: No bundle (reason: useless for push emails).