Added a test case for bug #9017.
item_cmpfunc.h:
A wrong not_null_tables method for Item_cond_xor
caused a conversion of a left join into an inner join
that was not valid.
and bug#8849 "problem with insert statement with table alias's":
make equality propagation work in stored procedures and prepared
statements.
Equality propagation can change AND/OR structure of ON expressions,
so the fix is to provide each execution of PS/SP with it's own
copy of AND/OR tree. We have been doing that already for WHERE clauses,
now ON clauses are also copied.
COND_EQUAL must be derived from Sql_alloc to simplify memory
management for objects of this class.
This fixes a leak in mysql-test-run noticed by PEM.
opt_range.cc:
Fixed uninitialized min_max_range member the QUICK_GROUP_MIN_MAX_SELECT class.
crashes server (prepared statements)": the bug was that all boolean
items always recovered its original arguments at statement cleanup
stage.
This collided with Item_subselect::select_transformer, which tries to
permanently change the item tree to use a transformed subselect instead of
original one.
So we had this call sequence for prepare:
mysql_stmt_prepare -> JOIN::prepare ->
Item_subselect::fix_fields -> the item tree gets transformed ->
Item_bool_rowready_func2::cleanup, item tree is recovered to original
state, while it shouldn't have been;
mysql_stmt_execute -> attempts to execute a broken tree -> crash.
Now instead of bluntly recovering all arguments of bool functions in
Item_bool_rowready_func2::cleanup, we recover only those
which were changed, and do it in one place.
There still would exist a possibility for a collision with subselect
tranformation, if permanent and temporary changes were performed at the
same stage.
But fortunately subselect transformation is always done first, so it
doesn't conflict with the optimization done by propogate_cond_constants.
Now we have:
mysql_stmt_prepare -> JOIN::prepare -> subselect transformation
permanently changes the tree -> cleanup doesn't recover anything,
because nothing was registered for recovery.
mysql_stmt_execute -> JOIN::prepare (the tree is already transformed,
so it doesn't change), JOIN::optimize ->
propogate_cond_constants -> temporary changes the item tree
with constants -> JOIN::execute -> cleanup ->
the changes done by propogate_cond_constants are recovered, as
they were registered for recovery.
The cause of the bug is that Item_func_in::fix_fields did not fully update its
used_table_cache. This was the cause for not_null_tables in setup_conds() to be
still 0 after the call
not_null_tables= (*conds)->not_null_tables();
As a result the condition in setup_conds()
if ( ... (table->table->map & not_null_tables) ...)
failed, which was the cause for the ON expression not to be added to conds, and later
the optimizer couldn't detect that it could apply the OUTER JOIN ==> JOIN optimization.
Added test case for bug #4256.
join_outer.result:
Fixed bug #4256.
item_cmpfunc.h:
Fixed inconsistency of values of used_tables_cache and
const_item_cache for Item_func_isnull objects.
This inconsistency caused bug #4256.
