bk commit into 5.0 tree (ted:1.2382) BUG#25219

[Vladimir Shebordaev]

Below is the list of changes that have just been committed into a local
5.0 repository of ted. When ted does a push these changes will
be propagated to the main repository and, within 24 hours after the
push, to the public repository.
For information on how to access the public repository
see http://dev.mysql.com/doc/mysql/en/installing-source-tree.html

ChangeSet@stripped, 2007-01-18 11:10:53+03:00, ted@stripped +1 -0
  BUG#25219: SELECT with subqueries causes MySQL server crash
  
  fix: subselect joins should be just partially cleaned up if they are known
       to be uncacheable

  sql/sql_select.cc@stripped, 2007-01-18 11:10:50+03:00, ted@stripped +3 -2
    JOIN::join_free() now does not want to purge uncacheable subselects

# This is a BitKeeper patch.  What follows are the unified diffs for the
# set of deltas contained in the patch.  The rest of the patch, the part
# that BitKeeper cares about, is below these diffs.
# User:	ted
# Host:	ted.mysql.internal
# Root:	/usr/local/src/mysql/bug25219/50

--- 1.478/sql/sql_select.cc	2007-01-18 11:10:59 +03:00
+++ 1.479/sql/sql_select.cc	2007-01-18 11:10:59 +03:00
@@ -6090,11 +6090,12 @@
 {
   SELECT_LEX_UNIT *unit;
   SELECT_LEX *sl;
+  Item_subselect *subselect = select_lex->master_unit()->item;
   /*
     Optimization: if not EXPLAIN and we are done with the JOIN,
     free all tables.
   */
-  bool full= (!select_lex->uncacheable && !thd->lex->describe);
+  bool full= ((!subselect || !subselect->is_uncacheable()) &&
!select_lex->uncacheable && !thd->lex->describe);
   bool can_unlock= full;
   DBUG_ENTER("JOIN::join_free");
 
@@ -6103,7 +6104,7 @@
   for (unit= select_lex->first_inner_unit(); unit; unit= unit->next_unit())
     for (sl= unit->first_select(); sl; sl= sl->next_select())
     {
-      Item_subselect *subselect= sl->master_unit()->item;
+      subselect= sl->master_unit()->item;
       bool full_local= full && (!subselect || subselect->is_evaluated());
       /*
         If this join is evaluated, we can fully clean it up and clean up all