with the test case added by this patch.
The bug cannot be reproduced with the same test case for the main
5.3 tree because the backported fix for bug 59696 masks the
problem that causes the crash in the mentioned test case. It's not
clear weather this fix masks this problem in all possible cases.
Anyway the patch for bug 698882 introduced some inconsistent data
structures that could contain indirect references to deleted object.
It happened when two Item_equal objects were merged and the Item_field
list of the second object was joined to such list of the first object.
This operation required adjustment of the backward pointers in
Item fields from the joined list. However the adjustment was missing
and this caused crashes in the tree for mwl#128.
Now the backward pointers are set only when Item_equal items are
completely built and are not changed anymore.
The patch fixed the following optimizer defect: when performing
substitution for best equal fields into where conditions to be
able to do their evaluations as soon as possible the optimizer
skipped conditions over views. That could lead to suboptimal
execution of queries that used views.
Slightly changed the test case to demonstrate the performance
improvements if this fix.
This bug could manifest itself when hash join over a varchar column
with NULL values in some rows was used. It happened because the
function key_buf_cmp erroneously returned FALSE when one of the joined
key fields was null while the second was not.
Also fixed two other bugs in the functions key_hashnr and key_buf_cmp
that could possibly lead to wrong results for some queries that
used hash join over several columns with nulls.
Also reverted the latest addition of the test case for bug #45092. It
had been already backported earlier.
This was another bug in the patch for bug 698882. The new
code from this patch did not ensured that substitutions
of fields for best equal fields were performed on all
AND-OR levels. As a result substitutions for best fields
in some predicates that had been used by the range optimizer
were not actually performed while range plans could employ
these substitutions. This could lead to inconsistent data
structures and ultimately to a crash.
The bug was in the code of the patch fixing bug 698882.
With improper casting the method store_key_field::change_source_field
was called for the elements of the array TABLE_REF::key_copy that
were either of a different type or not allocated at all. This caused
crashes in some queries.
Made sure that the optimal fields are used by TABLE_REF objects
when building index access keys to joined tables.
Fixed a bug in the template function that sorts the elements of
a list using the bubble sort algorithm. The bug caused poor
performance of the function. Also added an optimization that
skips comparison with the most heavy elements that has been
already properly placed in the list.
Made the comparison of the fields belonging to the same Item_equal
more granular: fields belonging to the same table are also ordered
according to some rules.
An assertion failure was triggered for a 6-way join query that used two
join buffers.
The failure happened because every call of JOIN_CACHE::join_matching_records
saved and restored status of all tables that were accessed before the table
join_tab. It must do it only for those of them that follow the last table
using a join buffer.
3150 Olav Sandstaa 2010-05-07
Test cases for the following "ICP for InnoDB" bugs:
Bug#40992 InnoDB: Crash when engine_condition_pushdown is on
Bug#35080 Innodb crash at mem_block_get_len line 72
Bug#41996 multi-table delete crashes server (InnoDB table)
Bug#43448 Server crashes on multi table delete with Innodb
All these bugs are duplicates of either one or both of Bug 43360 or 36981.
(backporting of olav@sun.com-20100226091930-qxvakxmcp6463t5w)
Date: Mon, 01 Nov 2010 15:15:25 -0000
3272 Roy Lyseng 2010-11-01
Bug#52068: Optimizer generates invalid semijoin materialization plan
When MaterializeScan semijoin strategy was used and there were one
or more outer dependent tables before the semijoin tables, the scan
over the materialized table was not properly reset for each row of
the prefix outer tables.
Example: suppose we have a join order:
ot1 SJ-Mat-Scan(it2 it3) ot4
Notice that this is called a MaterializeScan, even though there is an
outer table ahead of the materialized tables. Usually a MaterializeScan
has the outer tables after the materialized table, but this is
a special (but legal) case with outer dependent tables both before and
after the materialized table.
For each qualifying row from ot1, a new scan over the materialized
table must be set up. The code failed to do that, so all scans after
the first one returned zero rows from the materialized table.
- Cause: handler::in_range_check_pushed_down was not reset when a
command would call handler->idx_cond_push() without later calling
handler->index_end().
- Fix: reset the variable in handler->reset(), too (like we do with other
Index Condition Pushdown members).
One of the hash functions employed by the BNLH join algorithm
calculates the the value of hash index for key value utilizing
every byte of the key buffer. To make this calculation valid
one has to ensure that for any key value unused bytes of the
buffer are filled with with a certain filler. We choose 0 as
a filler for these bytes.
Added an optional boolean parameter with_zerofill to the function
key_copy. If the value of the parameter is TRUE all unused bytes
of the key buffer is filled with 0.
In some cases the function make_cond_for_index() was mistaken
when detecting index only pushdown conditions for a table:
a pushdown condition that was not index only could be marked
as such.
It happened because the procedure erroneously used the markers
for index only conditions that remained from the calls of
this function that extracted the index conditions for other
tables.
Fixed by erasing index only markers as soon as they are need
anymore.
- This allows us to detect missing my_afree() calls and also find overruns (when running with valgrind) to alloca() areas.
- Added missing my_afree() calls
- Fixed wrong call to my_afree()
include/my_sys.h:
When compiling with valgrind, change my_alloca() to use my_malloc()
mysql-test/suite/innodb/t/innodb_bug57255.test:
Speed up taste case (patch from Stewart Smith)
mysql-test/suite/innodb_plugin/t/innodb_bug57255.test:
Speed up taste case (patch from Stewart Smith)
sql/ha_partition.cc:
Removed casts from my_afree()
sql/opt_range.cc:
Add missing my_afree() calls.
storage/maria/ma_rt_split.c:
Fixed wrong parameter to my_afree()
