bk commit into 5.0 tree (konstantin:1.2007)

[konstantin]

Below is the list of changes that have just been committed into a local
5.0 repository of kostja. When kostja 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
  1.2007 05/06/24 21:06:59 konstantin@stripped +4 -0
  - don't call JOIN::join_free(1) twice for every join in JOIN::cleanup().
  The reason it happened was that both, JOIN::cleanup() and JOIN::join_free(),
  went over all nested joins and called cleanup/join_free for them.
  For that:
  - split recursive and non-recursive parts of JOIN::cleanup() and
  JOIN::join_free()
  - rename JOIN::cleanup to JOIN::destroy, as it actually destroys its
    argument
  - move the recursive part of JOIN::cleanup to st_select_lex::cleanup
  - move the non-recursive part of JOIN::join_free to the introduced
    method JOIN::cleanup().

  sql/sql_union.cc
    1.121 05/06/24 21:06:53 konstantin@stripped +29 -21
    Add st_select_lex::cleanup, a counterpart of st_select_lex_unit::cleanup():
    move the recursive part of JOIN::cleanup to it.

  sql/sql_select.h
    1.89 05/06/24 21:06:53 konstantin@stripped +9 -1
    - declarations for JOIN::cleanup() and JOIN::join_free()

  sql/sql_select.cc
    1.349 05/06/24 21:06:53 konstantin@stripped +70 -63
    - remove two unused arguments from return_zero_rows
    - split JOIN::join_free and JOIN::cleanup to recursive and non-recursive
      parts.
    - note, the assert in JOIN::join_free _does_ fail in having.test.
      We have two options: a) propagate `full' flag to the nested joins.
      We did it before, and this patch didn't change it. If so, we 
      can end up cleaning up an uncacheable JOIN (that is, the join that
      we might need again).
      b) evaluate own 'full' flag on every level. In this case, we might
      end up with tables freed in mysql_unlock_read_tables, but not
      cleaned up properly, and this may be even worse. The test suite
      passes with both approaches, but not with the assert.

  sql/sql_lex.h
    1.186 05/06/24 21:06:53 konstantin@stripped +5 -0
    Add st_select_lex::cleanup, a counterpart of st_select_lex_unit::cleanup()

# 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:	konstantin
# Host:	dragonfly.local
# Root:	/opt/local/work/mysql-5.0-join_free_anew

--- 1.185/sql/sql_lex.h	2005-06-21 18:18:23 +04:00
+++ 1.186/sql/sql_lex.h	2005-06-24 21:06:53 +04:00
@@ -642,6 +642,11 @@
   static void print_order(String *str, ORDER *order);
   void print_limit(THD *thd, String *str);
   void fix_prepare_information(THD *thd, Item **conds);
+  /*
+    Destroy the used execution plan (JOIN) of this subtree (this
+    SELECT_LEX and all nested SELECT_LEXes and SELECT_LEX_UNITs).
+  */
+  bool cleanup();
 };
 typedef class st_select_lex SELECT_LEX;
 

--- 1.348/sql/sql_select.cc	2005-06-23 20:22:02 +04:00
+++ 1.349/sql/sql_select.cc	2005-06-24 21:06:53 +04:00
@@ -87,10 +87,9 @@
 static ORDER *remove_const(JOIN *join,ORDER *first_order,COND *cond,
 			   bool change_list, bool *simple_order);
 static int return_zero_rows(JOIN *join, select_result *res,TABLE_LIST *tables,
-			    List<Item> &fields, bool send_row,
-			    uint select_options, const char *info,
-			    Item *having, Procedure *proc,
-			    SELECT_LEX_UNIT *unit);
+                            List<Item> &fields, bool send_row,
+                            uint select_options, const char *info,
+                            Item *having);
 static COND *build_equal_items(THD *thd, COND *cond,
                                COND_EQUAL *inherited,
                                List<TABLE_LIST> *join_list,
@@ -1227,8 +1226,7 @@
 			    send_row_on_empty_set(),
 			    select_options,
 			    zero_result_cause,
-			    having, procedure,
-			    unit);
+			    having);
     DBUG_VOID_RETURN;
   }
 
@@ -1437,7 +1435,7 @@
 	DBUG_VOID_RETURN;
       }
       end_read_record(&curr_join->join_tab->read_record);
-      curr_join->const_tables= curr_join->tables; // Mark free for join_free()
+      curr_join->const_tables= curr_join->tables; // Mark free for cleanup()
       curr_join->join_tab[0].table= 0;           // Table is freed
       
       // No sum funcs anymore
@@ -1667,9 +1665,9 @@
 */
 
 int
-JOIN::cleanup()
+JOIN::destroy()
 {
-  DBUG_ENTER("JOIN::cleanup");
+  DBUG_ENTER("JOIN::destroy");
   select_lex->join= 0;
 
   if (tmp_join)
@@ -1684,12 +1682,11 @@
     }
     tmp_join->tmp_join= 0;
     tmp_table_param.copy_field=0;
-    DBUG_RETURN(tmp_join->cleanup());
+    DBUG_RETURN(tmp_join->destroy());
   }
   cond_equal= 0;
 
-  lock=0;                                     // It's faster to unlock later
-  join_free(1);
+  cleanup(1);
   if (exec_tmp_table1)
     free_tmp_table(thd, exec_tmp_table1);
   if (exec_tmp_table2)
@@ -1697,12 +1694,6 @@
   delete select;
   delete_dynamic(&keyuse);
   delete procedure;
-  for (SELECT_LEX_UNIT *lex_unit= select_lex->first_inner_unit();
-       lex_unit != 0;
-       lex_unit= lex_unit->next_unit())
-  {
-    error|= lex_unit->cleanup();
-  }
   DBUG_RETURN(error);
 }
 
@@ -1885,17 +1876,10 @@
   THD *thd= join->thd;
   DBUG_ENTER("Cursor::close");
 
-  join->join_free(0);
   if (unit)
-  {
-    /* In case of UNIONs JOIN is freed inside unit->cleanup() */
-    unit->cleanup();
-  }
+    (void) unit->cleanup();
   else
-  {
-    join->cleanup();
-    delete join;
-  }
+    (void) join->select_lex->cleanup();
   {
     /* XXX: Another hack: closing tables used in the cursor */
     DBUG_ASSERT(lock || open_tables || derived_tables);
@@ -2071,8 +2055,7 @@
   if (free_join)
   {
     thd->proc_info="end";
-    err= join->cleanup();
-    delete join;
+    err= select_lex->cleanup();
     DBUG_RETURN(err || thd->net.report_error);
   }
   DBUG_RETURN(join->error);
@@ -5905,29 +5888,77 @@
 }
 
 
+void JOIN::join_free(bool full)
+{
+  SELECT_LEX_UNIT *unit;
+  SELECT_LEX *sl;
+  DBUG_ENTER("JOIN::join_free");
+  DBUG_PRINT("enter", ("this=0x%x, full=%d", this, (int) full));
+
+  /*
+    Optimization: if not EXPLAIN and we are done with the JOIN,
+    free all tables.
+  */
+  full= full || (!select_lex->uncacheable && !thd->lex->subqueries
&&
+                 !thd->lex->describe);
+
+  cleanup(full);
+
+  for (unit= select_lex->first_inner_unit(); unit; unit= unit->next_unit())
+    for (sl= unit->first_select_in_union(); sl; sl= sl->next_select())
+    {
+      JOIN *join= sl->join;
+      if (join)
+      {
+        /* Check that we don't occasionally clean up an uncacheable JOIN */
+#if 0
+        DBUG_ASSERT(! (!select_lex->uncacheable && sl->uncacheable));
+#endif
+        join->join_free(full);
+      }
+    }
+
+  /*
+    We are not using tables anymore
+    Unlock all tables. We may be in an INSERT .... SELECT statement.
+  */
+  if (full && lock && thd->lock && !(select_options &
SELECT_NO_UNLOCK) &&
+      !select_lex->subquery_in_having &&
+      (select_lex == (thd->lex->unit.fake_select_lex ?
+                      thd->lex->unit.fake_select_lex :
&thd->lex->select_lex)))
+  {
+    /*
+      TODO: unlock tables even if the join isn't top level select in the
+      tree.
+    */
+    mysql_unlock_read_tables(thd, lock);           // Don't free join->lock
+    lock= 0;
+  }
+
+  DBUG_VOID_RETURN;
+}
+
+
 /*
   Free resources of given join
 
   SYNOPSIS
-    JOIN::join_free()
+    JOIN::cleanup()
     fill - true if we should free all resources, call with full==1 should be
            last, before it this function can be called with full==0
 
   NOTE: with subquery this function definitely will be called several times,
     but even for simple query it can be called several times.
 */
-void
-JOIN::join_free(bool full)
-{
-  JOIN_TAB *tab,*end;
-  DBUG_ENTER("JOIN::join_free");
 
-  full= full || (!select_lex->uncacheable &&
-                 !thd->lex->subqueries &&
-                 !thd->lex->describe); // do not cleanup too early on EXPLAIN
+void JOIN::cleanup(bool full)
+{
+  DBUG_ENTER("JOIN::cleanup");
+  DBUG_PRINT("enter", ("this=0x%x, full=%d\n", this, (int) full));
 
   if (table)
   {
+    JOIN_TAB *tab,*end;
     /*
       Only a sorted table may be cached.  This sorted table is always the
       first non const table in join->table
@@ -5938,16 +5969,6 @@
       filesort_free_buffers(table[const_tables]);
     }
 
-    for (SELECT_LEX_UNIT *unit= select_lex->first_inner_unit(); unit;
-         unit= unit->next_unit())
-    {
-      JOIN *join;
-      for (SELECT_LEX *sl= unit->first_select_in_union(); sl;
-           sl= sl->next_select())
-        if ((join= sl->join))
-          join->join_free(full);
-    }
-
     if (full)
     {
       for (tab= join_tab, end= tab+tables; tab != end; tab++)
@@ -5964,23 +5985,10 @@
       }
     }
   }
-
   /*
     We are not using tables anymore
     Unlock all tables. We may be in an INSERT .... SELECT statement.
   */
-  if (full && lock && thd->lock && !(select_options &
SELECT_NO_UNLOCK) &&
-      !select_lex->subquery_in_having)
-  {
-    // TODO: unlock tables even if the join isn't top level select in the tree
-    if (select_lex == (thd->lex->unit.fake_select_lex ?
-                       thd->lex->unit.fake_select_lex :
&thd->lex->select_lex))
-    {
-      mysql_unlock_read_tables(thd, lock);        // Don't free join->lock
-      lock=0;
-    }
-  }
-
   if (full)
   {
     group_fields.delete_elements();
@@ -6217,8 +6225,7 @@
 static int
 return_zero_rows(JOIN *join, select_result *result,TABLE_LIST *tables,
 		 List<Item> &fields, bool send_row, uint select_options,
-		 const char *info, Item *having, Procedure *procedure,
-		 SELECT_LEX_UNIT *unit)
+		 const char *info, Item *having)
 {
   DBUG_ENTER("return_zero_rows");
 

--- 1.88/sql/sql_select.h	2005-06-22 11:59:07 +04:00
+++ 1.89/sql/sql_select.h	2005-06-24 21:06:53 +04:00
@@ -325,7 +325,7 @@
   int optimize();
   int reinit();
   void exec();
-  int cleanup();
+  int destroy();
   void restore_tmp();
   bool alloc_func_list();
   bool make_sum_func_list(List<Item> &all_fields, List<Item>
&send_fields,
@@ -349,7 +349,15 @@
   int rollup_send_data(uint idx);
   int rollup_write_data(uint idx, TABLE *table);
   bool test_in_subselect(Item **where);
+  /*
+    Release memory and, if possible, the open tables held by this execution
+    plan (and nested plans). It's used to release some tables before
+    the end of execution in order to increase concurrency and reduce
+    memory consumption.
+  */
   void join_free(bool full);
+  /* Cleanup this JOIN, possibly for reuse */
+  void cleanup(bool full);
   void clear();
   bool save_join_tab();
   bool send_row_on_empty_set()

--- 1.120/sql/sql_union.cc	2005-06-15 21:58:23 +04:00
+++ 1.121/sql/sql_union.cc	2005-06-24 21:06:53 +04:00
@@ -553,7 +553,6 @@
 bool st_select_lex_unit::cleanup()
 {
   int error= 0;
-  JOIN *join;
   DBUG_ENTER("st_select_lex_unit::cleanup");
 
   if (cleaned)
@@ -572,29 +571,17 @@
   }
 
   for (SELECT_LEX *sl= first_select_in_union(); sl; sl= sl->next_select())
+    error|= sl->cleanup();
+
+  if (fake_select_lex)
   {
-    if ((join= sl->join))
-    {
-      error|= sl->join->cleanup();
-      delete join;
-    }
-    else
+    JOIN *join;
+    if ((join= fake_select_lex->join))
     {
-      // it can be DO/SET with subqueries
-      for (SELECT_LEX_UNIT *lex_unit= sl->first_inner_unit();
-	   lex_unit != 0;
-	   lex_unit= lex_unit->next_unit())
-      {
-	error|= lex_unit->cleanup();
-      }
+      join->tables_list= 0;
+      join->tables= 0;
     }
-  }
-  if (fake_select_lex && (join= fake_select_lex->join))
-  {
-    join->tables_list= 0;
-    join->tables= 0;
-    error|= join->cleanup();
-    delete join;
+    error|= fake_select_lex->cleanup();
   }
 
   DBUG_RETURN(error);
@@ -650,3 +637,24 @@
     res= fake_select_lex->join->change_result(result);
   return (res);
 }
+
+
+bool st_select_lex::cleanup()
+{
+  bool error= FALSE;
+  DBUG_ENTER("st_select_lex::cleanup()");
+
+  if (join)
+  {
+    error|= join->destroy();
+    delete join;
+    join= 0;
+  }
+  for (SELECT_LEX_UNIT *lex_unit= first_inner_unit(); lex_unit ;
+       lex_unit= lex_unit->next_unit())
+  {
+    error|= lex_unit->cleanup();
+  }
+  DBUG_RETURN(error);
+}
+