Merge spetrunia@bk-internal.mysql.com:/home/bk/mysql-5.0-opt

into  mysql.com:/home/psergey/mysql-5.0-bug8804-r12

mysql-test/r/ndb_subquery.result

@@ -9,7 +9,7 @@ insert into t2 values (1,1,1),(2,2,2),(3,3,3), (4,4,4), (5,5,5);
 explain select * from t2 where p NOT IN (select p from t1);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	5	Using where
-2	DEPENDENT SUBQUERY	t1	unique_subquery	PRIMARY	PRIMARY	4	func	1	Using index
+2	DEPENDENT SUBQUERY	t1	unique_subquery	PRIMARY	PRIMARY	4	func	1	
 select * from t2 where p NOT IN (select p from t1) order by p;
 p	u	o
 4	4	4
@@ -17,7 +17,7 @@ p	u	o
 explain select * from t2 where p NOT IN (select u from t1);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	5	Using where
-2	DEPENDENT SUBQUERY	t1	unique_subquery	u	u	4	func	1	Using index
+2	DEPENDENT SUBQUERY	t1	unique_subquery	u	u	4	func	1	
 select * from t2 where p NOT IN (select u from t1) order by p;
 p	u	o
 4	4	4
@@ -25,7 +25,7 @@ p	u	o
 explain select * from t2 where p NOT IN (select o from t1);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	NULL	NULL	NULL	NULL	5	Using where
-2	DEPENDENT SUBQUERY	t1	index_subquery	o	o	4	func	1	Using index
+2	DEPENDENT SUBQUERY	t1	index_subquery	o	o	4	func	1	
 select * from t2 where p NOT IN (select o from t1) order by p;
 p	u	o
 4	4	4

mysql-test/r/subselect2.result

@@ -126,11 +126,11 @@ id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t2	ALL	DDOCTYPEID_IDX	NULL	NULL	NULL	9	Using where
 1	PRIMARY	t1	eq_ref	PRIMARY	PRIMARY	34	test.t2.DOCID	1	
 1	PRIMARY	t4	eq_ref	PRIMARY	PRIMARY	34	test.t2.DOCTYPEID	1	
-2	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using index; Using where
-3	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using index; Using where
-4	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using index; Using where
-5	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using index; Using where
-6	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX,CMFLDRPARNT_IDX	PRIMARY	34	func	1	Using index; Using where
+2	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using where
+3	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using where
+4	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using where
+5	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX	PRIMARY	34	func	1	Using where
+6	DEPENDENT SUBQUERY	t3	unique_subquery	PRIMARY,FFOLDERID_IDX,CMFLDRPARNT_IDX	PRIMARY	34	func	1	Using where
 drop table t1, t2, t3, t4;
 CREATE TABLE t1 (a int(10) , PRIMARY KEY (a)) Engine=InnoDB;
 INSERT INTO t1 VALUES (1),(2);

sql/item_cmpfunc.cc

@@ -829,11 +829,35 @@ longlong Item_in_optimizer::val_int()
           We disable the predicates we've pushed down into subselect, run the
           subselect and see if it has produced any rows.
         */
-        ((Item_in_subselect*)args[1])->enable_pushed_conds= FALSE;
-        longlong tmp= args[1]->val_bool_result();
-        result_for_null_param= null_value= 
-          !((Item_in_subselect*)args[1])->engine->no_rows();
-        ((Item_in_subselect*)args[1])->enable_pushed_conds= TRUE;
+        Item_in_subselect *item_subs=(Item_in_subselect*)args[1]; 
+        if (cache->cols() == 1)
+        {
+          item_subs->set_cond_guard_var(0, FALSE);
+          longlong tmp= args[1]->val_bool_result();
+          result_for_null_param= null_value= !item_subs->engine->no_rows();
+          item_subs->set_cond_guard_var(0, TRUE);
+        }
+        else
+        {
+          uint i;
+          uint ncols= cache->cols();
+          /*
+            Turn off the predicates that are based on column compares for
+            which the left part is currently NULL
+          */
+          for (i= 0; i < ncols; i++)
+          {
+            if (cache->el(i)->null_value)
+              item_subs->set_cond_guard_var(i, FALSE);
+          }
+          
+          longlong tmp= args[1]->val_bool_result();
+          result_for_null_param= null_value= !item_subs->engine->no_rows();
+          
+          /* Turn all predicates back on */
+          for (i= 0; i < ncols; i++)
+            item_subs->set_cond_guard_var(i, TRUE);
+        }
       }
     }
     return 0;

sql/item_cmpfunc.h

@@ -312,6 +312,7 @@ public:
   enum Functype functype() const { return TRIG_COND_FUNC; };
   const char *func_name() const { return "trigcond"; };
   bool const_item() const { return FALSE; }
+  bool *get_trig_var() { return trig_var; }
 };
 
 class Item_func_not_all :public Item_func_not

sql/item_subselect.h

@@ -94,7 +94,7 @@ public:
     return null_value;
   }
   bool fix_fields(THD *thd, Item **ref);
-  virtual bool exec(bool full_scan);
+  virtual bool exec();
   virtual void fix_length_and_dec();
   table_map used_tables() const;
   table_map not_null_tables() const { return 0; }
@@ -104,6 +104,7 @@ public:
   Item *get_tmp_table_item(THD *thd);
   void update_used_tables();
   void print(String *str);
+  virtual bool have_guarded_conds() { return FALSE; }
   bool change_engine(subselect_engine *eng)
   {
     old_engine= engine;
@@ -249,13 +250,21 @@ protected:
   bool transformed;
 public:
   /* Used to trigger on/off conditions that were pushed down to subselect */
-  bool enable_pushed_conds;
+  bool *pushed_cond_guards;
+
+  bool *get_cond_guard(int i)
+  {
+    return pushed_cond_guards ? pushed_cond_guards + i : NULL;
+  }
+  void set_cond_guard_var(int i, bool v) { pushed_cond_guards[i]= v; }
+  bool have_guarded_conds() { return test(pushed_cond_guards); }
+
   Item_func_not_all *upper_item; // point on NOT/NOP before ALL/SOME subquery
 
   Item_in_subselect(Item * left_expr, st_select_lex *select_lex);
   Item_in_subselect()
     :Item_exists_subselect(), optimizer(0), abort_on_null(0), transformed(0),
-     enable_pushed_conds(TRUE), upper_item(0)
+     pushed_cond_guards(NULL), upper_item(0)
   {}
 
   subs_type substype() { return IN_SUBS; }
@@ -340,23 +349,22 @@ public:
 
     SYNOPSIS
       exec()
-        full_scan TRUE  - Pushed-down predicates are disabled, the engine
-                          must disable made based on those predicates.
-                  FALSE - Pushed-down predicates are in effect.
+
     DESCRIPTION
       Execute the engine. The result of execution is subquery value that is
       either captured by previously set up select_result-based 'sink' or
       stored somewhere by the exec() method itself.
 
-      A required side effect: if full_scan==TRUE, subselect_engine->no_rows()
-      should return correct result.
+      A required side effect: If at least one pushed-down predicate is
+      disabled, subselect_engine->no_rows() must return correct result after 
+      the exec() call.
 
     RETURN
       0 - OK
-      1 - Either an execution error, or the engine was be "changed", and
+      1 - Either an execution error, or the engine was "changed", and the
           caller should call exec() again for the new engine.
   */
-  virtual int exec(bool full_scan)= 0;
+  virtual int exec()= 0;
   virtual uint cols()= 0; /* return number of columns in select */
   virtual uint8 uncacheable()= 0; /* query is uncacheable */
   enum Item_result type() { return res_type; }
@@ -391,7 +399,7 @@ public:
   void cleanup();
   int prepare();
   void fix_length_and_dec(Item_cache** row);
-  int exec(bool full_scan);
+  int exec();
   uint cols();
   uint8 uncacheable();
   void exclude();
@@ -415,7 +423,7 @@ public:
   void cleanup();
   int prepare();
   void fix_length_and_dec(Item_cache** row);
-  int exec(bool full_scan);
+  int exec();
   uint cols();
   uint8 uncacheable();
   void exclude();
@@ -429,11 +437,30 @@ public:
 
 
 struct st_join_table;
+
+
+/*
+  A subquery execution engine that evaluates the subquery by doing one index
+  lookup in a unique index.
+
+  This engine is used to resolve subqueries in forms
+  
+    outer_expr IN (SELECT tbl.unique_key FROM tbl WHERE subq_where) 
+    
+  or, tuple-based:
+  
+    (oe1, .. oeN) IN (SELECT uniq_key_part1, ... uniq_key_partK
+                      FROM tbl WHERE subqwhere) 
+  
+  i.e. the subquery is a single table SELECT without GROUP BY, aggregate
+  functions, etc.
+*/
+
 class subselect_uniquesubquery_engine: public subselect_engine
 {
 protected:
   st_join_table *tab;
-  Item *cond;
+  Item *cond; /* The WHERE condition of subselect */
   /* 
     TRUE<=> last execution produced empty set. Valid only when left
     expression is NULL.
@@ -453,7 +480,7 @@ public:
   void cleanup();
   int prepare();
   void fix_length_and_dec(Item_cache** row);
-  int exec(bool full_scan);
+  int exec();
   uint cols() { return 1; }
   uint8 uncacheable() { return UNCACHEABLE_DEPENDENT; }
   void exclude();
@@ -471,16 +498,47 @@ class subselect_indexsubquery_engine: public subselect_uniquesubquery_engine
 {
   /* FALSE for 'ref', TRUE for 'ref-or-null'. */
   bool check_null;
+  /* 
+    The "having" clause. This clause (further reffered to as "artificial
+    having") was inserted by subquery transformation code. It contains 
+    Item(s) that have a side-effect: they record whether the subquery has 
+    produced a row with NULL certain components. We need to use it for cases
+    like
+      (oe1, oe2) IN (SELECT t.key, t.no_key FROM t1)
+    where we do index lookup on t.key=oe1 but need also to check if there
+    was a row such that t.no_key IS NULL.
+    
+    NOTE: This is currently here and not in the uniquesubquery_engine. Ideally
+    it should have been in uniquesubquery_engine in order to allow execution of
+    subqueries like
+    
+      (oe1, oe2) IN (SELECT primary_key, non_key_maybe_null_field FROM tbl)
+
+    We could use uniquesubquery_engine for the first component and let
+    Item_is_not_null_test( non_key_maybe_null_field) to handle the second.
+
+    However, subqueries like the above are currently not handled by index
+    lookup-based subquery engines, the engine applicability check misses
+    them: it doesn't switch the engine for case of artificial having and
+    [eq_]ref access (only for artifical having + ref_or_null or no having).
+    The above example subquery is handled as a full-blown SELECT with eq_ref
+    access to one table.
+
+    Due to this limitation, the "artificial having" currently needs to be 
+    checked by only in indexsubquery_engine.
+  */
+  Item *having;
 public:
 
   // constructor can assign THD because it will be called after JOIN::prepare
   subselect_indexsubquery_engine(THD *thd, st_join_table *tab_arg,
 				 Item_subselect *subs, Item *where,
-				 bool chk_null)
+                                 Item *having_arg, bool chk_null)
     :subselect_uniquesubquery_engine(thd, tab_arg, subs, where),
-     check_null(chk_null)
+     check_null(chk_null),
+     having(having_arg)
   {}
-  int exec(bool full_scan);
+  int exec();
   void print (String *str);
 };
 

sql/mysql_priv.h

@@ -1201,7 +1201,7 @@ extern const char *command_name[];
 extern const char *first_keyword, *my_localhost, *delayed_user, *binary_keyword;
 extern const char **errmesg;			/* Error messages */
 extern const char *myisam_recover_options_str;
-extern const char *in_left_expr_name, *in_additional_cond;
+extern const char *in_left_expr_name, *in_additional_cond, *in_having_cond;
 extern const char * const triggers_file_ext;
 extern const char * const trigname_file_ext;
 extern Eq_creator eq_creator;

sql/mysqld.cc

@@ -453,10 +453,13 @@ char *mysqld_unix_port, *opt_mysql_tmpdir;
 const char **errmesg;			/* Error messages */
 const char *myisam_recover_options_str="OFF";
 const char *myisam_stats_method_str="nulls_unequal";
+
 /* name of reference on left espression in rewritten IN subquery */
 const char *in_left_expr_name= "<left expr>";
 /* name of additional condition */
 const char *in_additional_cond= "<IN COND>";
+const char *in_having_cond= "<IN HAVING>";
+
 my_decimal decimal_zero;
 /* classes for comparation parsing/processing */
 Eq_creator eq_creator;

sql/sql_lex.h

@@ -469,7 +469,7 @@ public:
   void set_thd(THD *thd_arg) { thd= thd_arg; }
 
   friend void lex_start(THD *thd, uchar *buf, uint length);
-  friend int subselect_union_engine::exec(bool);
+  friend int subselect_union_engine::exec();
 
   List<Item> *get_unit_column_types();
 };

sql/sql_select.h

@@ -35,8 +35,17 @@ typedef struct keyuse_t {
     satisfied if val has NULL 'value'.
   */
   bool null_rejecting;
-  /* TRUE<=> This ref access is an outer subquery reference access */
-  bool outer_ref;
+  /*
+    !NULL - This KEYUSE was created from an equality that was wrapped into
+            an Item_func_trig_cond. This means the equality (and validity of 
+            this KEYUSE element) can be turned on and off. The on/off state 
+            is indicted by the pointed value:
+              *cond_guard == TRUE <=> equality condition is on
+              *cond_guard == FALSE <=> equality condition is off
+
+    NULL  - Otherwise (the source equality can't be turned off)
+  */
+  bool *cond_guard;
 } KEYUSE;
 
 class store_key;
@@ -51,6 +60,18 @@ typedef struct st_table_ref
   byte          *key_buff2;               // key_buff+key_length
   store_key     **key_copy;               //
   Item          **items;                  // val()'s for each keypart
+  /*  
+    Array of pointers to trigger variables. Some/all of the pointers may be
+    NULL.  The ref access can be used iff
+    
+      for each used key part i, (!cond_guards[i] || *cond_guards[i]) 
+
+    This array is used by subquery code. The subquery code may inject
+    triggered conditions, i.e. conditions that can be 'switched off'. A ref 
+    access created from such condition is not valid when at least one of the 
+    underlying conditions is switched off (see subquery code for more details)
+  */
+  bool          **cond_guards;
   /*
     (null_rejecting & (1<<i)) means the condition is '=' and no matching
     rows will be produced if items[i] IS NULL (see add_not_null_conds())
@@ -99,6 +120,13 @@ enum enum_nested_loop_state
   NESTED_LOOP_QUERY_LIMIT= 3, NESTED_LOOP_CURSOR_LIMIT= 4
 };
 
+
+/* Values for JOIN_TAB::packed_info */
+#define TAB_INFO_HAVE_VALUE 1
+#define TAB_INFO_USING_INDEX 2
+#define TAB_INFO_USING_WHERE 4
+#define TAB_INFO_FULL_SCAN_ON_NULL 8
+
 typedef enum_nested_loop_state
 (*Next_select_func)(JOIN *, struct st_join_table *, bool);
 typedef int (*Read_record_func)(struct st_join_table *tab);
@@ -119,7 +147,15 @@ typedef struct st_join_table {
   st_join_table *last_inner;    /* last table table for embedding outer join */
   st_join_table *first_upper;  /* first inner table for embedding outer join */
   st_join_table *first_unmatched; /* used for optimization purposes only     */
+  
+  /* Special content for EXPLAIN 'Extra' column or NULL if none */
   const char	*info;
+  /* 
+    Bitmap of TAB_INFO_* bits that encodes special line for EXPLAIN 'Extra'
+    column, or 0 if there is no info.
+  */
+  uint          packed_info;
+
   Read_record_func read_first_record;
   Next_select_func next_select;
   READ_RECORD	read_record;
@@ -386,7 +422,7 @@ public:
 			  Item_sum ***func);
   int rollup_send_data(uint idx);
   int rollup_write_data(uint idx, TABLE *table);
-  bool test_in_subselect(Item **where);
+  void remove_subq_pushed_predicates(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