Merge bb-10.1-explain-analyze into 10.1

mysql-test/r/analyze_stmt.result

@@ -0,0 +1,149 @@
+drop table if exists t0,t1,t2,t3;
+create table t0 (a int) engine=myisam;
+INSERT INTO t0 VALUES (0),(1),(2),(3),(4),(5),(6),(7),(8),(9);
+create table t1 (a int) engine=myisam;
+INSERT INTO t1 select * from t0;
+# Try a few basic selects to see that r_rows and r_filtered columns work
+analyze select * from t1;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	100.00	
+analyze select * from t1 where a<5;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where
+analyze select * from t1 where a>100;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	0.00	Using where
+# ANALYZE DELETE will delete rows:
+analyze delete from t1 where a in (2,3,4);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	NULL	100.00	30.00	Using where
+select * from t1;
+a
+0
+1
+5
+6
+7
+8
+9
+drop table t1;
+# ANALYZE UPDATE will make updates:
+create table t1(a int, b int);
+insert into t1 select a,a from t0;
+analyze update t1 set b=100+b where a in (6,7,8);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	NULL	100.00	30.00	Using where
+select * from t1;
+a	b
+0	0
+1	1
+2	2
+3	3
+4	4
+5	5
+6	106
+7	107
+8	108
+9	9
+drop table t1;
+# Check that UNION works
+create table t1(a int, b int);
+insert into t1 select a,a from t0;
+analyze (select * from t1 A where a<5) union (select * from t1 B where a in (5,6));
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	PRIMARY	A	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where
+2	UNION	B	ALL	NULL	NULL	NULL	NULL	10	10	100.00	20.00	Using where
+NULL	UNION RESULT	<union1,2>	ALL	NULL	NULL	NULL	NULL	NULL	7	NULL	NULL	
+analyze (select * from t1 A where a<5) union (select * from t1 B where a in (1,2));
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	PRIMARY	A	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where
+2	UNION	B	ALL	NULL	NULL	NULL	NULL	10	10	100.00	20.00	Using where
+NULL	UNION RESULT	<union1,2>	ALL	NULL	NULL	NULL	NULL	NULL	5	NULL	NULL	
+drop table t1;
+drop table t0;
+# 
+# Try a subquery.
+# 
+create table t0 (a int, b int);
+insert into t0 values
+(0,0),(1,1),(2,2),(3,3),(4,4),(5,5),(6,6),(7,7),(8,8),(9,9);
+create table t1 (a int, b int);
+insert into t1 values (1,1),(2,2),(3,3);
+# See .test file for the right values of r_rows and r_filtered.
+analyze select a, a in (select t0.b from t0 where t0.b+1=t1.b+1) from t1;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	3	3	100.00	100.00	
+2	DEPENDENT SUBQUERY	t0	ALL	NULL	NULL	NULL	NULL	10	3	100.00	33.33	Using where
+# Try a subquery that is never executed
+analyze select a, a in (select t0.b from t0 where t0.b+1=t1.b+1) from t1 where t1.a > 5;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	3	3	100.00	0.00	Using where
+2	DEPENDENT SUBQUERY	t0	ALL	NULL	NULL	NULL	NULL	10	NULL	100.00	NULL	Using where
+drop table t0, t1;
+#
+# Tests for join buffering
+#
+create table t0 (a int, b int);
+insert into t0 values
+(0,0),(1,1),(2,2),(3,3),(4,4),(5,5),(6,6),(7,7),(8,8),(9,9);
+create table t1 like t0;
+insert into t1 select * from t0;
+explain select * from t0, t1 where t0.a<5 and t1.a<5;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	SIMPLE	t0	ALL	NULL	NULL	NULL	NULL	10	Using where
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	Using where; Using join buffer (flat, BNL join)
+# These should have filtered=50
+analyze select * from t0, t1 where t0.a<5 and t1.a<5;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t0	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where; Using join buffer (flat, BNL join)
+explain select * from t0, t1 where t0.a<5 and t1.b=t0.b;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	SIMPLE	t0	ALL	NULL	NULL	NULL	NULL	10	Using where
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	Using where; Using join buffer (flat, BNL join)
+# Now, t1 should have filtered=10
+analyze select * from t0, t1 where t0.a<5 and t1.b=t0.b;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t0	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	10.00	Using where; Using join buffer (flat, BNL join)
+explain select * from t0, t1 where t0.a<5 and t1.a<5 and t1.b=t0.b;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	SIMPLE	t0	ALL	NULL	NULL	NULL	NULL	10	Using where
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	Using where; Using join buffer (flat, BNL join)
+# Now, t1 should have filtered=10
+analyze select * from t0, t1 where t0.a<5 and t1.a<5 and t1.b=t0.b;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t0	ALL	NULL	NULL	NULL	NULL	10	10	100.00	50.00	Using where
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	10.00	Using where; Using join buffer (flat, BNL join)
+# TODO: Check what is counted for "range checked for each record".
+#
+# Test for joins
+#
+create table t2 (key1 int, key2x int, col1 int, key(key1), key(key2x));
+insert into t2 select A.a + 10 *B.a +100 * C.a, 
+(A.a + 10 *B.a +100 * C.a)*2,
+A.a + 10 *B.a +100 * C.a 
+from t0 A, t0 B, t0 C;
+# This always has matches, filtered=100%.
+analyze select * from t1,t2 where t2.key1=t1.a;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	100.00	Using where
+1	SIMPLE	t2	ref	key1	key1	5	test.t1.a	1	1	100.00	100.00	
+# This shows r_rows=0. It is actually 0.5 (should r_rows be changed to double?)
+analyze select * from t1,t2 where t2.key2x=t1.a;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	100.00	Using where
+1	SIMPLE	t2	ref	key2x	key2x	5	test.t1.a	1	0	100.00	100.00	
+select * from t1,t2 where t2.key2x=t1.a;
+a	b	key1	key2x	col1
+0	0	0	0	0
+2	2	1	2	1
+4	4	2	4	2
+6	6	3	6	3
+8	8	4	8	4
+# This has t2.filtered=40% (there are 5 values: {0,1,2,3,4}. two of them have mod=0)
+analyze select * from t1,t2 where t2.key2x=t1.a and mod(t2.col1,4)=0;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	r_rows	filtered	r_filtered	Extra
+1	SIMPLE	t1	ALL	NULL	NULL	NULL	NULL	10	10	100.00	100.00	Using where
+1	SIMPLE	t2	ref	key2x	key2x	5	test.t1.a	1	0	100.00	40.00	Using where
+drop table t0,t1,t2;

mysql-test/r/ps.result

@@ -1,5 +1,6 @@
 call mtr.add_suppression('Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT.');
 drop table if exists t1,t2,t3,t4;
+drop database if exists mysqltest1;
 drop database if exists client_test_db;
 create table t1
 (

mysql-test/r/show_explain.result

@@ -641,7 +641,7 @@ set debug_dbug='+d,show_explain_probe_join_exec_start';
 SHOW INDEX FROM t1;
 show explain for $thr2;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
-1	SIMPLE	STATISTICS	ALL	NULL	NULL	NULL	NULL	NULL	Skip_open_table; Scanned all databases
+1	SIMPLE	STATISTICS	ALL	NULL	TABLE_SCHEMA,TABLE_NAME	NULL	NULL	NULL	Open_full_table; Scanned 0 databases
 Warnings:
 Note	1003	SHOW INDEX FROM t1
 Table	Non_unique	Key_name	Seq_in_index	Column_name	Collation	Cardinality	Sub_part	Packed	Null	Index_type	Comment	Index_comment

mysql-test/t/analyze_stmt.test

@@ -0,0 +1,109 @@
+#
+#  Tests for "ANALYZE $statement" feature (PostgreSQL's analog is called EXPLAIN ANALYZE)
+#
+--disable_warnings
+drop table if exists t0,t1,t2,t3;
+--enable_warnings
+
+create table t0 (a int) engine=myisam;
+INSERT INTO t0 VALUES (0),(1),(2),(3),(4),(5),(6),(7),(8),(9);
+
+create table t1 (a int) engine=myisam;
+INSERT INTO t1 select * from t0;
+
+--echo # Try a few basic selects to see that r_rows and r_filtered columns work
+analyze select * from t1;
+analyze select * from t1 where a<5;
+analyze select * from t1 where a>100;
+
+--echo # ANALYZE DELETE will delete rows:
+analyze delete from t1 where a in (2,3,4);
+select * from t1;
+drop table t1;
+
+--echo # ANALYZE UPDATE will make updates:
+create table t1(a int, b int);
+insert into t1 select a,a from t0;
+analyze update t1 set b=100+b where a in (6,7,8);
+select * from t1;
+drop table t1;
+
+--echo # Check that UNION works
+create table t1(a int, b int);
+insert into t1 select a,a from t0;
+analyze (select * from t1 A where a<5) union (select * from t1 B where a in (5,6));
+analyze (select * from t1 A where a<5) union (select * from t1 B where a in (1,2));
+drop table t1;
+drop table t0;
+
+--echo # 
+--echo # Try a subquery.
+--echo # 
+create table t0 (a int, b int);
+insert into t0 values
+  (0,0),(1,1),(2,2),(3,3),(4,4),(5,5),(6,6),(7,7),(8,8),(9,9);
+
+create table t1 (a int, b int);
+insert into t1 values (1,1),(2,2),(3,3);
+
+#
+#   t1      t0
+#   a=1     (0,1)       2 rows
+#   a=2     (0,1,2)     3 rows
+#   a=3     (0,1,2,3)   4 rows
+#
+#  TOTAL    TOTAL= 9 rows.  3 executions, avg=3 rows.
+#  WHERE is satisfied for 1 row per query, which gives filtered=33.3
+
+--echo # See .test file for the right values of r_rows and r_filtered.
+analyze select a, a in (select t0.b from t0 where t0.b+1=t1.b+1) from t1;
+
+--echo # Try a subquery that is never executed
+analyze select a, a in (select t0.b from t0 where t0.b+1=t1.b+1) from t1 where t1.a > 5;
+
+drop table t0, t1;
+
+--echo #
+--echo # Tests for join buffering
+--echo #
+create table t0 (a int, b int);
+insert into t0 values
+  (0,0),(1,1),(2,2),(3,3),(4,4),(5,5),(6,6),(7,7),(8,8),(9,9);
+create table t1 like t0;
+insert into t1 select * from t0;
+
+explain select * from t0, t1 where t0.a<5 and t1.a<5;
+--echo # These should have filtered=50
+analyze select * from t0, t1 where t0.a<5 and t1.a<5;
+
+explain select * from t0, t1 where t0.a<5 and t1.b=t0.b;
+--echo # Now, t1 should have filtered=10
+analyze select * from t0, t1 where t0.a<5 and t1.b=t0.b;
+
+explain select * from t0, t1 where t0.a<5 and t1.a<5 and t1.b=t0.b;
+--echo # Now, t1 should have filtered=10
+analyze select * from t0, t1 where t0.a<5 and t1.a<5 and t1.b=t0.b;
+
+--echo # TODO: Check what is counted for "range checked for each record".
+
+--echo #
+--echo # Test for joins
+--echo #
+create table t2 (key1 int, key2x int, col1 int, key(key1), key(key2x));
+insert into t2 select A.a + 10 *B.a +100 * C.a, 
+                      (A.a + 10 *B.a +100 * C.a)*2,
+                      A.a + 10 *B.a +100 * C.a 
+               from t0 A, t0 B, t0 C;
+
+--echo # This always has matches, filtered=100%.
+analyze select * from t1,t2 where t2.key1=t1.a;
+
+--echo # This shows r_rows=0. It is actually 0.5 (should r_rows be changed to double?)
+analyze select * from t1,t2 where t2.key2x=t1.a;
+        select * from t1,t2 where t2.key2x=t1.a;
+
+--echo # This has t2.filtered=40% (there are 5 values: {0,1,2,3,4}. two of them have mod=0)
+analyze select * from t1,t2 where t2.key2x=t1.a and mod(t2.col1,4)=0;
+
+drop table t0,t1,t2;
+

mysql-test/t/ps.test

@@ -7,6 +7,7 @@ call mtr.add_suppression('Unsafe statement written to the binary log using state
 --disable_warnings
 drop table if exists t1,t2,t3,t4;
 
+drop database if exists mysqltest1;
 # Avoid wrong warnings if mysql_client_test fails
 drop database if exists client_test_db;
 --enable_warnings

sql/protocol.h

@@ -210,6 +210,42 @@ public:
   virtual enum enum_protocol_type type() { return PROTOCOL_BINARY; };
 };
 
+
+/*
+  A helper for "ANALYZE $stmt" which looks a real network procotol but doesn't
+  write results to the network.
+
+  At first glance, class select_send looks like a more appropriate place to
+  implement the "write nothing" hook. This is not true, because
+    - we need to evaluate the value of every item, and do it the way
+      select_send does it (i.e. call item->val_int() or val_real() or...)
+    - select_send::send_data() has some other code, like telling the storage
+      engine that the row can be unlocked. We want to keep that also.
+  as a result, "ANALYZE $stmt" uses a select_send_analyze which still uses 
+  select_send::send_data() & co., and also uses  Protocol_discard object.
+*/
+
+class Protocol_discard : public Protocol_text
+{
+public:
+  Protocol_discard(THD *thd_arg) : Protocol_text(thd_arg) {}
+  /* The real writing is done only in write() */
+  virtual bool write() { return 0; }
+  virtual bool send_result_set_metadata(List<Item> *list, uint flags)
+  {
+    // Don't pas Protocol::SEND_NUM_ROWS | Protocol::SEND_EOF flags 
+    return Protocol_text::send_result_set_metadata(list, 0);
+  }
+
+  // send_error is intentionally not overloaded.
+  virtual bool send_eof(uint server_status, uint statement_warn_count)
+  {
+    return 0;
+  }
+
+};
+
+
 void send_warning(THD *thd, uint sql_errno, const char *err=0);
 bool net_send_error(THD *thd, uint sql_errno, const char *err,
                     const char* sqlstate);

sql/sql_class.cc

@@ -2282,6 +2282,9 @@ int THD::send_explain_fields(select_result *result)
 /*
   Populate the provided field_list with EXPLAIN output columns.
   this->lex->describe has the EXPLAIN flags
+
+  The set/order of columns must be kept in sync with 
+  Explain_query::print_explain and co.
 */
 
 void THD::make_explain_field_list(List<Item> &field_list)
@@ -2317,11 +2320,25 @@ void THD::make_explain_field_list(List<Item> &field_list)
   item->maybe_null=1;
   field_list.push_back(item= new Item_return_int("rows", 10,
                                                  MYSQL_TYPE_LONGLONG));
-  if (lex->describe & DESCRIBE_EXTENDED)
+  if (lex->analyze_stmt)
+  {
+    field_list.push_back(item= new Item_return_int("r_rows", 10,
+                                                   MYSQL_TYPE_LONGLONG));
+    item->maybe_null=1;
+  }
+
+  if (lex->analyze_stmt || lex->describe & DESCRIBE_EXTENDED)
   {
     field_list.push_back(item= new Item_float("filtered", 0.1234, 2, 4));
     item->maybe_null=1;
   }
+
+  if (lex->analyze_stmt)
+  {
+    field_list.push_back(item= new Item_float("r_filtered", 0.1234, 2, 4));
+    item->maybe_null=1;
+  }
+
   item->maybe_null= 1;
   field_list.push_back(new Item_empty_string("Extra", 255, cs));
 }

sql/sql_class.h

@@ -3961,6 +3961,20 @@ public:
 };
 
 
+/*
+  We need this class, because select_send::send_eof() will call ::my_eof.
+
+  See also class Protocol_discard.
+*/
+
+class select_send_analyze : public select_send
+{
+  bool send_result_set_metadata(List<Item> &list, uint flags) { return 0; }
+  bool send_eof() { return 0; }
+  void abort_result_set() {}
+};
+
+
 class select_to_file :public select_result_interceptor {
 protected:
   sql_exchange *exchange;

sql/sql_delete.cc

@@ -223,6 +223,7 @@ bool mysql_delete(THD *thd, TABLE_LIST *table_list, COND *conds,
   killed_state killed_status= NOT_KILLED;
   THD::enum_binlog_query_type query_type= THD::ROW_QUERY_TYPE;
   bool with_select= !select_lex->item_list.is_empty();
+  Explain_delete *explain;
   Delete_plan query_plan(thd->mem_root);
   query_plan.index= MAX_KEY;
   query_plan.using_filesort= FALSE;
@@ -538,9 +539,11 @@ bool mysql_delete(THD *thd, TABLE_LIST *table_list, COND *conds,
       goto cleanup;
   }
 
+  explain= (Explain_delete*)thd->lex->explain->get_upd_del_plan();
   while (!(error=info.read_record(&info)) && !thd->killed &&
 	 ! thd->is_error())
   {
+    explain->on_record_read();
     if (table->vfield)
       update_virtual_fields(thd, table,
                             table->triggers ? VCOL_UPDATE_ALL :
@@ -549,6 +552,7 @@ bool mysql_delete(THD *thd, TABLE_LIST *table_list, COND *conds,
     // thd->is_error() is tested to disallow delete row on error
     if (!select || select->skip_record(thd) > 0)
     {
+      explain->on_record_after_where();
       if (table->triggers &&
           table->triggers->process_triggers(thd, TRG_EVENT_DELETE,
                                             TRG_ACTION_BEFORE, FALSE))
@@ -666,6 +670,11 @@ cleanup:
   }
   DBUG_ASSERT(transactional_table || !deleted || thd->transaction.stmt.modified_non_trans_table);
   free_underlaid_joins(thd, select_lex);
+  if (thd->lex->analyze_stmt)
+  {
+    error= thd->lex->explain->send_explain(thd);
+  }
+  else
   if (error < 0 || 
       (thd->lex->ignore && !thd->is_error() && !thd->is_fatal_error))
   {

sql/sql_explain.cc

@@ -140,7 +140,7 @@ int Explain_query::send_explain(THD *thd)
     return 1;
 
   int res;
-  if ((res= print_explain(result, lex->describe)))
+  if ((res= print_explain(result, lex->describe, lex->analyze_stmt)))
     result->abort_result_set();
   else
     result->send_eof();
@@ -154,16 +154,16 @@ int Explain_query::send_explain(THD *thd)
 */
 
 int Explain_query::print_explain(select_result_sink *output, 
-                                 uint8 explain_flags)
+                                 uint8 explain_flags, bool is_analyze)
 {
   if (upd_del_plan)
   {
-    upd_del_plan->print_explain(this, output, explain_flags);
+    upd_del_plan->print_explain(this, output, explain_flags, is_analyze);
     return 0;
   }
   else if (insert_plan)
   {
-    insert_plan->print_explain(this, output, explain_flags);
+    insert_plan->print_explain(this, output, explain_flags, is_analyze);
     return 0;
   }
   else
@@ -172,14 +172,14 @@ int Explain_query::print_explain(select_result_sink *output,
     Explain_node *node= get_node(1);
     if (!node)
       return 1; /* No query plan */
-    return node->print_explain(this, output, explain_flags);
+    return node->print_explain(this, output, explain_flags, is_analyze);
   }
 }
 
 
 bool print_explain_query(LEX *lex, THD *thd, String *str)
 {
-  return lex->explain->print_explain_str(thd, str);
+  return lex->explain->print_explain_str(thd, str, false);
 }
 
 
@@ -187,14 +187,14 @@ bool print_explain_query(LEX *lex, THD *thd, String *str)
   Return tabular EXPLAIN output as a text string
 */
 
-bool Explain_query::print_explain_str(THD *thd, String *out_str)
+bool Explain_query::print_explain_str(THD *thd, String *out_str,  bool is_analyze)
 {
   List<Item> fields;
   thd->make_explain_field_list(fields);
 
   select_result_text_buffer output_buf(thd);
   output_buf.send_result_set_metadata(fields, thd->lex->describe);
-  if (print_explain(&output_buf, thd->lex->describe))
+  if (print_explain(&output_buf, thd->lex->describe, is_analyze))
     return true;
   output_buf.save_to(out_str);
   return false;
@@ -217,7 +217,8 @@ static void push_string(List<Item> *item_list, String *str)
 
 int Explain_union::print_explain(Explain_query *query, 
                                  select_result_sink *output,
-                                 uint8 explain_flags)
+                                 uint8 explain_flags, 
+                                 bool is_analyze)
 {
   char table_name_buffer[SAFE_NAME_LEN];
 
@@ -225,7 +226,7 @@ int Explain_union::print_explain(Explain_query *query,
   for (int i= 0; i < (int) union_members.elements(); i++)
   {
     Explain_select *sel= query->get_select(union_members.at(i));
-    sel->print_explain(query, output, explain_flags);
+    sel->print_explain(query, output, explain_flags, is_analyze);
   }
 
   /* Print a line with "UNION RESULT" */
@@ -287,9 +288,21 @@ int Explain_union::print_explain(Explain_query *query,
  
   /* `rows` */
   item_list.push_back(item_null);
+  
+  /* `r_rows` */
+  if (is_analyze)
+  {
+    ha_rows avg_rows= fake_select_lex_tracker.get_avg_rows();
+    item_list.push_back(new Item_int((longlong) (ulonglong) avg_rows,
+                                      MY_INT64_NUM_DECIMAL_DIGITS));
+  }
 
   /* `filtered` */
-  if (explain_flags & DESCRIBE_EXTENDED)
+  if (explain_flags & DESCRIBE_EXTENDED || is_analyze)
+    item_list.push_back(item_null);
+
+  /* `r_filtered` */
+  if (is_analyze)
     item_list.push_back(item_null);
 
   /* `Extra` */
@@ -309,7 +322,7 @@ int Explain_union::print_explain(Explain_query *query,
     Print all subquery children (UNION children have already been printed at
     the start of this function)
   */
-  return print_explain_for_children(query, output, explain_flags);
+  return print_explain_for_children(query, output, explain_flags, is_analyze);
 }
 
 
@@ -319,18 +332,26 @@ int Explain_union::print_explain(Explain_query *query,
 
 int Explain_node::print_explain_for_children(Explain_query *query, 
                                          select_result_sink *output,
-                                         uint8 explain_flags)
+                                         uint8 explain_flags, 
+                                         bool is_analyze)
 {
   for (int i= 0; i < (int) children.elements(); i++)
   {
     Explain_node *node= query->get_node(children.at(i));
-    if (node->print_explain(query, output, explain_flags))
+    if (node->print_explain(query, output, explain_flags, is_analyze))
       return 1;
   }
   return 0;
 }
 
 
+void Explain_select::replace_table(uint idx, Explain_table_access *new_tab)
+{
+  delete join_tabs[idx];
+  join_tabs[idx]= new_tab;
+}
+
+
 Explain_select::~Explain_select()
 {
   if (join_tabs)
@@ -344,7 +365,7 @@ Explain_select::~Explain_select()
 
 int Explain_select::print_explain(Explain_query *query, 
                                   select_result_sink *output,
-                                  uint8 explain_flags)
+                                  uint8 explain_flags, bool is_analyze)
 {
   if (message)
   {
@@ -359,8 +380,17 @@ int Explain_select::print_explain(Explain_query *query,
       item_list.push_back(item_null);
     if (explain_flags & DESCRIBE_PARTITIONS)
       item_list.push_back(item_null);
-    if (explain_flags & DESCRIBE_EXTENDED)
+
+    /* filtered */
+    if (is_analyze || explain_flags & DESCRIBE_EXTENDED)
       item_list.push_back(item_null);
+    
+    if (is_analyze)
+    {
+      /* r_rows, r_filtered */
+      item_list.push_back(item_null);
+      item_list.push_back(item_null);
+    }
 
     item_list.push_back(new Item_string(message,strlen(message),cs));
 
@@ -373,7 +403,7 @@ int Explain_select::print_explain(Explain_query *query,
     bool using_fs= using_filesort;
     for (uint i=0; i< n_join_tabs; i++)
     {
-      join_tabs[i]->print_explain(output, explain_flags, select_id,
+      join_tabs[i]->print_explain(output, explain_flags, is_analyze, select_id,
                                   select_type, using_tmp, using_fs);
       if (i == 0)
       {
@@ -387,7 +417,7 @@ int Explain_select::print_explain(Explain_query *query,
     }
   }
 
-  return print_explain_for_children(query, output, explain_flags);
+  return print_explain_for_children(query, output, explain_flags, is_analyze);
 }
 
 
@@ -398,8 +428,9 @@ void Explain_table_access::push_extra(enum explain_extra_tag extra_tag)
 
 
 int Explain_table_access::print_explain(select_result_sink *output, uint8 explain_flags, 
-                                    uint select_id, const char *select_type,
+                                        bool is_analyze,
-                                    bool using_temporary, bool using_filesort)
+                                        uint select_id, const char *select_type,
+                                        bool using_temporary, bool using_filesort)
 {
   const CHARSET_INFO *cs= system_charset_info;
   const char *hash_key_prefix= "#hash#";
@@ -519,8 +550,23 @@ int Explain_table_access::print_explain(select_result_sink *output, uint8 explai
   else
     item_list.push_back(item_null);
 
+  /* `r_rows` */
+  if (is_analyze)
+  {
+    if (!tracker.has_scans())
+    {
+      item_list.push_back(item_null);
+    }
+    else
+    {
+      ha_rows avg_rows= tracker.get_avg_rows();
+      item_list.push_back(new Item_int((longlong) (ulonglong) avg_rows,
+                                        MY_INT64_NUM_DECIMAL_DIGITS));
+    }
+  }
+
   /* `filtered` */
-  if (explain_flags & DESCRIBE_EXTENDED)
+  if (explain_flags & DESCRIBE_EXTENDED || is_analyze)
   {
     if (filtered_set)
     {
@@ -530,6 +576,22 @@ int Explain_table_access::print_explain(select_result_sink *output, uint8 explai
       item_list.push_back(item_null);
   }
 
+  /* `r_filtered` */
+  if (is_analyze)
+  {
+    if (!tracker.has_scans())
+    {
+      item_list.push_back(item_null);
+    }
+    else
+    {
+      double r_filtered= tracker.get_filtered_after_where();
+      if (bka_type.is_using_jbuf())
+        r_filtered *= jbuf_tracker.get_filtered_after_where();
+      item_list.push_back(new Item_float(r_filtered*100.0, 2));
+    }
+  }
+
   /* `Extra` */
   StringBuffer<256> extra_buf;
   bool first= true;
@@ -802,12 +864,13 @@ void Explain_quick_select::print_key_len(String *str)
 
 int Explain_delete::print_explain(Explain_query *query, 
                                   select_result_sink *output,
-                                  uint8 explain_flags)
+                                  uint8 explain_flags,
+                                  bool is_analyze)
 {
   if (deleting_all_rows)
   {
     const char *msg= "Deleting all rows";
-    int res= print_explain_message_line(output, explain_flags,
+    int res= print_explain_message_line(output, explain_flags, is_analyze,
                                         1 /*select number*/,
                                         select_type, &rows, msg);
     return res;
@@ -815,14 +878,16 @@ int Explain_delete::print_explain(Explain_query *query,
   }
   else
   {
-    return Explain_update::print_explain(query, output, explain_flags);
+    return Explain_update::print_explain(query, output, explain_flags,
+                                         is_analyze);
   }
 }
 
 
 int Explain_update::print_explain(Explain_query *query, 
                                   select_result_sink *output,
-                                  uint8 explain_flags)
+                                  uint8 explain_flags,
+                                  bool is_analyze)
 {
   StringBuffer<64> key_buf;
   StringBuffer<64> key_len_buf;
@@ -832,7 +897,7 @@ int Explain_update::print_explain(Explain_query *query,
     const char *msg= impossible_where ? 
                      "Impossible WHERE" : 
                      "No matching rows after partition pruning";
-    int res= print_explain_message_line(output, explain_flags,
+    int res= print_explain_message_line(output, explain_flags, is_analyze,
                                         1 /*select number*/,
                                         select_type, 
                                         NULL, /* rows */
@@ -892,8 +957,9 @@ int Explain_update::print_explain(Explain_query *query,
     Single-table DELETE commands do not do "Using temporary".
     "Using index condition" is also not possible (which is an unjustified limitation)
   */
+  double r_filtered= 100 * (r_rows?((double)r_rows_after_where/r_rows):1.0); 
 
-  print_explain_row(output, explain_flags, 
+  print_explain_row(output, explain_flags, is_analyze,
                     1, /* id */
                     select_type,
                     table_name.c_ptr(), 
@@ -904,18 +970,21 @@ int Explain_update::print_explain(Explain_query *query,
                     key_len_buf.length() ? key_len_buf.c_ptr() : NULL,
                     NULL, /* 'ref' is always NULL in single-table EXPLAIN DELETE */
                     &rows,
+                    &r_rows,
+                    r_filtered,
                     extra_str.c_ptr_safe());
 
-  return print_explain_for_children(query, output, explain_flags);
+  return print_explain_for_children(query, output, explain_flags, is_analyze);
 }
 
 
 int Explain_insert::print_explain(Explain_query *query, 
                                   select_result_sink *output, 
-                                  uint8 explain_flags)
+                                  uint8 explain_flags,
+                                  bool is_analyze)
 {
   const char *select_type="INSERT";
-  print_explain_row(output, explain_flags, 
+  print_explain_row(output, explain_flags, is_analyze,
                     1, /* id */
                     select_type,
                     table_name.c_ptr(), 
@@ -926,9 +995,11 @@ int Explain_insert::print_explain(Explain_query *query,
                     NULL, // key_len
                     NULL, // ref
                     NULL, // rows
+                    NULL, // r_rows
+                    100.0, // r_filtered
                     NULL);
 
-  return print_explain_for_children(query, output, explain_flags);
+  return print_explain_for_children(query, output, explain_flags, is_analyze);
 }
 
 

sql/sql_explain.h

@@ -14,6 +14,38 @@
    along with this program; if not, write to the Free Software
    Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA */
 
+/* Data structures for ANALYZE */
+class Table_access_tracker 
+{
+public:
+  Table_access_tracker() :
+    r_scans(0), r_rows(0), /*r_rows_after_table_cond(0),*/
+    r_rows_after_where(0)
+  {}
+
+  ha_rows r_scans; /* How many scans were ran on this join_tab */
+  ha_rows r_rows; /* How many rows we've got after that */
+//  ha_rows r_rows_after_table_cond; /* Rows after applying the table condition */
+  ha_rows r_rows_after_where; /* Rows after applying attached part of WHERE */
+
+  bool has_scans() { return (r_scans != 0); }
+  ha_rows get_avg_rows()
+  {
+    return r_scans ? (ha_rows)rint((double) r_rows / r_scans): 0;
+  }
+
+  double get_filtered_after_where()
+  {
+    double r_filtered;
+    if (r_rows > 0)
+      r_filtered= (double)r_rows_after_where / r_rows;
+    else
+      r_filtered= 1.0;
+
+    return r_filtered;
+  }
+};
+
 
 /**************************************************************************************
  
@@ -60,10 +92,10 @@ public:
   }
 
   virtual int print_explain(Explain_query *query, select_result_sink *output, 
-                            uint8 explain_flags)=0;
+                            uint8 explain_flags, bool is_analyze)=0;
   
   int print_explain_for_children(Explain_query *query, select_result_sink *output, 
-                                 uint8 explain_flags);
+                                 uint8 explain_flags, bool is_analyze);
   virtual ~Explain_node(){}
 };
 
@@ -109,6 +141,12 @@ public:
     join_tabs[n_join_tabs++]= tab;
     return false;
   }
+  
+  /*
+    This is used to save the results of "late" test_if_skip_sort_order() calls
+    that are made from JOIN::exec
+  */
+  void replace_table(uint idx, Explain_table_access *new_tab);
 
 public:
   int select_id;
@@ -134,7 +172,14 @@ public:
   bool using_filesort;
   
   int print_explain(Explain_query *query, select_result_sink *output, 
-                    uint8 explain_flags);
+                    uint8 explain_flags, bool is_analyze);
+  
+  Table_access_tracker *get_using_temporary_read_tracker()
+  {
+    return &using_temporary_read_tracker;
+  }
+private:
+  Table_access_tracker using_temporary_read_tracker;
 };
 
 
@@ -172,10 +217,23 @@ public:
     union_members.append(select_no);
   }
   int print_explain(Explain_query *query, select_result_sink *output, 
-                    uint8 explain_flags);
+                    uint8 explain_flags, bool is_analyze);
 
   const char *fake_select_type;
   bool using_filesort;
+
+  Table_access_tracker *get_fake_select_lex_tracker()
+  {
+    return &fake_select_lex_tracker;
+  }
+  Table_access_tracker *get_tmptable_read_tracker()
+  {
+    return &tmptable_read_tracker;
+  }
+private:
+  Table_access_tracker fake_select_lex_tracker;
+  /* This one is for reading after ORDER BY */
+  Table_access_tracker tmptable_read_tracker; 
 };
 
 
@@ -183,6 +241,7 @@ class Explain_update;
 class Explain_delete;
 class Explain_insert;
 
+
 /*
   Explain structure for a query (i.e. a statement).
 
@@ -238,13 +297,14 @@ public:
   Explain_union *get_union(uint select_id);
  
   /* Produce a tabular EXPLAIN output */
-  int print_explain(select_result_sink *output, uint8 explain_flags);
+  int print_explain(select_result_sink *output, uint8 explain_flags, 
+                    bool is_analyze);
   
   /* Send tabular EXPLAIN to the client */
   int send_explain(THD *thd);
   
   /* Return tabular EXPLAIN output as a text string */
-  bool print_explain_str(THD *thd, String *out_str);
+  bool print_explain_str(THD *thd, String *out_str, bool is_analyze);
 
   /* If true, at least part of EXPLAIN can be printed */
   bool have_query_plan() { return insert_plan || upd_del_plan|| get_node(1) != NULL; }
@@ -252,6 +312,8 @@ public:
   void query_plan_ready();
 
   MEM_ROOT *mem_root;
+
+  Explain_update *get_upd_del_plan() { return upd_del_plan; }
 private:
   /* Explain_delete inherits from Explain_update */
   Explain_update *upd_del_plan;
@@ -320,13 +382,17 @@ enum explain_extra_tag
 };
 
 
-typedef struct st_explain_bka_type
+class EXPLAIN_BKA_TYPE
 {
+public:
+  EXPLAIN_BKA_TYPE() : join_alg(NULL) {}
+
   bool incremental;
   const char *join_alg;
   StringBuffer<64> mrr_type;
-
+  
-} EXPLAIN_BKA_TYPE;
+  bool is_using_jbuf() { return (join_alg != NULL); }
+};
 
 
 /*
@@ -386,6 +452,7 @@ private:
 /*
   EXPLAIN data structure for a single JOIN_TAB.
 */
+
 class Explain_table_access : public Sql_alloc
 {
 public:
@@ -459,8 +526,14 @@ public:
   StringBuffer<32> firstmatch_table_name;
 
   int print_explain(select_result_sink *output, uint8 explain_flags, 
+                    bool is_analyze,
                     uint select_id, const char *select_type,
                     bool using_temporary, bool using_filesort);
+
+  /* ANALYZE members*/
+  Table_access_tracker tracker;
+  Table_access_tracker jbuf_tracker;
+
 private:
   void append_tag_name(String *str, enum explain_extra_tag tag);
 };
@@ -502,8 +575,17 @@ public:
   bool using_filesort;
   bool using_io_buffer;
 
+  /* ANALYZE members and methods */
+  ha_rows r_rows;
+  ha_rows r_rows_after_where;
+  inline void on_record_read() { r_rows++; }
+  inline void on_record_after_where() { r_rows_after_where++; }
+
+  Explain_update() : 
+    r_rows(0), r_rows_after_where(0)
+  {}
   virtual int print_explain(Explain_query *query, select_result_sink *output, 
-                            uint8 explain_flags);
+                            uint8 explain_flags, bool is_analyze);
 };
 
 
@@ -523,7 +605,7 @@ public:
   int get_select_id() { return 1; /* always root */ }
 
   int print_explain(Explain_query *query, select_result_sink *output, 
-                    uint8 explain_flags);
+                    uint8 explain_flags, bool is_analyze);
 };
 
 
@@ -544,7 +626,7 @@ public:
   virtual int get_select_id() { return 1; /* always root */ }
 
   virtual int print_explain(Explain_query *query, select_result_sink *output, 
-                            uint8 explain_flags);
+                            uint8 explain_flags, bool is_analyze);
 };
 
 

sql/sql_join_cache.cc

@@ -2260,7 +2260,7 @@ enum_nested_loop_state JOIN_CACHE::join_matching_records(bool skip_last)
     */
     goto finish;
   }
-  
+
   while (!(error= join_tab_scan->next()))   
   {
     if (join->thd->check_killed())
@@ -2277,11 +2277,13 @@ enum_nested_loop_state JOIN_CACHE::join_matching_records(bool skip_last)
     /* Prepare to read matching candidates from the join buffer */
     if (prepare_look_for_matches(skip_last))
       continue;
+    join_tab->jbuf_tracker->r_scans++;
 
     uchar *rec_ptr;
     /* Read each possible candidate from the buffer and look for matches */
     while ((rec_ptr= get_next_candidate_for_match()))
-    { 
+    {
+      join_tab->jbuf_tracker->r_rows++;
       /* 
         If only the first match is needed, and, it has been already found for
         the next record read from the join buffer, then the record is skipped.
@@ -2451,6 +2453,8 @@ inline bool JOIN_CACHE::check_match(uchar *rec_ptr)
 
   if (join_tab->select && join_tab->select->skip_record(join->thd) <= 0)
     DBUG_RETURN(FALSE);
+  
+  join_tab->jbuf_tracker->r_rows_after_where++;
 
   if (!join_tab->is_last_inner_table())
     DBUG_RETURN(TRUE);
@@ -2574,7 +2578,7 @@ finish:
     none
 */ 
 
-void JOIN_CACHE::save_explain_data(struct st_explain_bka_type *explain)
+void JOIN_CACHE::save_explain_data(EXPLAIN_BKA_TYPE *explain)
 {
   explain->incremental= MY_TEST(prev_cache);
 
@@ -2619,14 +2623,14 @@ static void add_mrr_explain_info(String *str, uint mrr_mode, handler *file)
   }
 }
 
-void JOIN_CACHE_BKA::save_explain_data(struct st_explain_bka_type *explain)
+void JOIN_CACHE_BKA::save_explain_data(EXPLAIN_BKA_TYPE *explain)
 {
   JOIN_CACHE::save_explain_data(explain); 
   add_mrr_explain_info(&explain->mrr_type, mrr_mode, join_tab->table->file);
 }
 
 
-void JOIN_CACHE_BKAH::save_explain_data(struct st_explain_bka_type *explain)
+void JOIN_CACHE_BKAH::save_explain_data(EXPLAIN_BKA_TYPE *explain)
 {
   JOIN_CACHE::save_explain_data(explain); 
   add_mrr_explain_info(&explain->mrr_type, mrr_mode, join_tab->table->file);
@@ -3333,6 +3337,7 @@ int JOIN_TAB_SCAN::open()
 {
   save_or_restore_used_tabs(join_tab, FALSE);
   is_first_record= TRUE;
+  join_tab->tracker->r_scans++;
   return join_init_read_record(join_tab);
 }
 
@@ -3371,8 +3376,14 @@ int JOIN_TAB_SCAN::next()
     is_first_record= FALSE;
   else
     err= info->read_record(info);
-  if (!err && table->vfield)
+
-    update_virtual_fields(thd, table);
+  if (!err)
+  {
+    join_tab->tracker->r_rows++;
+    if (table->vfield)
+      update_virtual_fields(thd, table);
+  }
+
   while (!err && select && (skip_rc= select->skip_record(thd)) <= 0)
   {
     if (thd->check_killed() || skip_rc < 0) 
@@ -3382,9 +3393,16 @@ int JOIN_TAB_SCAN::next()
       meet the condition pushed to the table join_tab.
     */
     err= info->read_record(info);
-    if (!err && table->vfield)
+    if (!err)
-      update_virtual_fields(thd, table);
+    {
-  } 
+      join_tab->tracker->r_rows++;
+      if (table->vfield)
+        update_virtual_fields(thd, table);
+    }
+  }
+
+  if (!err)
+    join_tab->tracker->r_rows_after_where++;
   return err; 
 }
 

sql/sql_join_cache.h

@@ -63,7 +63,7 @@ typedef struct st_cache_field {
 
 class JOIN_TAB_SCAN;
 
-struct st_explain_bka_type;
+class EXPLAIN_BKA_TYPE;
 
 /*
   JOIN_CACHE is the base class to support the implementations of 
@@ -662,7 +662,7 @@ public:
   enum_nested_loop_state join_records(bool skip_last);
 
   /* Add a comment on the join algorithm employed by the join cache */
-  virtual void save_explain_data(struct st_explain_bka_type *explain);
+  virtual void save_explain_data(EXPLAIN_BKA_TYPE *explain);
 
   THD *thd();
 
@@ -1340,7 +1340,7 @@ public:
   /* Check index condition of the joined table for a record from BKA cache */
   bool skip_index_tuple(range_id_t range_info);
 
-  void save_explain_data(struct st_explain_bka_type *explain);
+  void save_explain_data(EXPLAIN_BKA_TYPE *explain);
 };
 
 
@@ -1431,5 +1431,5 @@ public:
   /* Check index condition of the joined table for a record from BKAH cache */
   bool skip_index_tuple(range_id_t range_info);
 
-  void save_explain_data(struct st_explain_bka_type *explain);
+  void save_explain_data(EXPLAIN_BKA_TYPE *explain);
 };

sql/sql_lex.cc

@@ -483,6 +483,7 @@ void lex_start(THD *thd)
   if (lex->select_lex.group_list_ptrs)
     lex->select_lex.group_list_ptrs->clear();
   lex->describe= 0;
+  lex->analyze_stmt= 0;
   lex->subqueries= FALSE;
   lex->context_analysis_only= 0;
   lex->derived_tables= 0;
@@ -4181,12 +4182,12 @@ bool st_select_lex::is_merged_child_of(st_select_lex *ancestor)
 */
 
 int LEX::print_explain(select_result_sink *output, uint8 explain_flags,
-                       bool *printed_anything)
+                       bool is_analyze, bool *printed_anything)
 {
   int res;
   if (explain && explain->have_query_plan())
   {
-    res= explain->print_explain(output, explain_flags);
+    res= explain->print_explain(output, explain_flags, is_analyze);
     *printed_anything= true;
   }
   else

sql/sql_lex.h

@@ -2268,6 +2268,7 @@ public:
 
   void save_explain_data(Explain_query *query);
   void save_explain_data_intern(Explain_query *query, Explain_update *eu);
+
   virtual ~Update_plan() {}
 
   Update_plan(MEM_ROOT *mem_root_arg) : 
@@ -2459,6 +2460,7 @@ struct LEX: public Query_tables_list
   */
   uint table_count;
   uint8 describe;
+  bool  analyze_stmt; /* TRUE<=> this is "ANALYZE $stmt" */
   /*
     A flag that indicates what kinds of derived tables are present in the
     query (0 if no derived tables, otherwise a combination of flags
@@ -2735,7 +2737,7 @@ struct LEX: public Query_tables_list
   }
 
   int print_explain(select_result_sink *output, uint8 explain_flags,
-                    bool *printed_anything);
+                    bool is_analyze, bool *printed_anything);
 };
 
 

sql/sql_parse.cc

@@ -5233,7 +5233,8 @@ static bool execute_sqlcom_select(THD *thd, TABLE_LIST *all_tables)
           top-level LIMIT
         */        
         result->reset_offset_limit(); 
-        thd->lex->explain->print_explain(result, thd->lex->describe);
+        thd->lex->explain->print_explain(result, thd->lex->describe,
+                                         thd->lex->analyze_stmt);
         if (lex->describe & DESCRIBE_EXTENDED)
         {
           char buff[1024];
@@ -5257,12 +5258,37 @@ static bool execute_sqlcom_select(THD *thd, TABLE_LIST *all_tables)
     }
     else
     {
-      if (!result && !(result= new select_send()))
+      //psergey-todo: ANALYZE should hook in here...
-        return 1;                               /* purecov: inspected */
+      select_result *save_result;
+      Protocol *save_protocol;
+      if (lex->analyze_stmt)
+      {
+        save_result= result;
+        result= new select_send_analyze();
+        save_protocol= thd->protocol;
+        thd->protocol= new Protocol_discard(thd);
+      }
+      else
+      {
+        if (!result && !(result= new select_send()))
+          return 1;                               /* purecov: inspected */
+      }
       query_cache_store_query(thd, all_tables);
       res= handle_select(thd, lex, result, 0);
       if (result != lex->result)
         delete result;
+
+      if (lex->analyze_stmt)
+      {
+        result= save_result;
+        if (!result && !(result= new select_send()))
+          return 1;
+        thd->protocol= save_protocol;
+        thd->lex->explain->send_explain(thd);
+
+        if (result != lex->result)
+          delete result;
+      }
     }
   }
   /* Count number of empty select queries */

sql/sql_select.h

@@ -250,7 +250,9 @@ typedef struct st_join_table {
   
   /* Special content for EXPLAIN 'Extra' column or NULL if none */
   enum explain_extra_tag info;
-
+  
+  Table_access_tracker *tracker;
+  Table_access_tracker *jbuf_tracker;
   /* 
     Bitmap of TAB_INFO_* bits that encodes special line for EXPLAIN 'Extra'
     column, or 0 if there is no info.
@@ -536,6 +538,11 @@ typedef struct st_join_table {
   }
 
   void remove_redundant_bnl_scan_conds();
+
+  void save_explain_data(Explain_table_access *eta, table_map prefix_tables, 
+                         bool distinct, struct st_join_table *first_top_tab);
+
+  void update_explain_data(uint idx);
 } JOIN_TAB;
 
 
@@ -1848,14 +1855,14 @@ void push_index_cond(JOIN_TAB *tab, uint keyno);
 
 /* EXPLAIN-related utility functions */
 int print_explain_message_line(select_result_sink *result, 
-                               uint8 options,
+                               uint8 options, bool is_analyze,
                                uint select_number,
                                const char *select_type,
                                ha_rows *rows,
                                const char *message);
 void explain_append_mrr_info(QUICK_RANGE_SELECT *quick, String *res);
 int print_explain_row(select_result_sink *result,
-                      uint8 options,
+                      uint8 options, bool is_analyze,
                       uint select_number,
                       const char *select_type,
                       const char *table_name,
@@ -1866,6 +1873,8 @@ int print_explain_row(select_result_sink *result,
                       const char *key_len,
                       const char *ref,
                       ha_rows *rows,
+                      ha_rows *r_rows,
+                      double r_filtered,
                       const char *extra);
 void make_possible_keys_line(TABLE *table, key_map possible_keys, String *line);
 

sql/sql_show.cc

@@ -121,41 +121,6 @@ append_algorithm(TABLE_LIST *table, String *buff);
 
 static COND * make_cond_for_info_schema(COND *cond, TABLE_LIST *table);
 
-/**
-  Condition pushdown used for INFORMATION_SCHEMA / SHOW queries.
-  This structure is to implement an optimization when
-  accessing data dictionary data in the INFORMATION_SCHEMA
-  or SHOW commands.
-  When the query contain a TABLE_SCHEMA or TABLE_NAME clause,
-  narrow the search for data based on the constraints given.
-*/
-typedef struct st_lookup_field_values
-{
-  /**
-    Value of a TABLE_SCHEMA clause.
-    Note that this value length may exceed @c NAME_LEN.
-    @sa wild_db_value
-  */
-  LEX_STRING db_value;
-  /**
-    Value of a TABLE_NAME clause.
-    Note that this value length may exceed @c NAME_LEN.
-    @sa wild_table_value
-  */
-  LEX_STRING table_value;
-  /**
-    True when @c db_value is a LIKE clause,
-    false when @c db_value is an '=' clause.
-  */
-  bool wild_db_value;
-  /**
-    True when @c table_value is a LIKE clause,
-    false when @c table_value is an '=' clause.
-  */
-  bool wild_table_value;
-} LOOKUP_FIELD_VALUES;
-
-
 bool get_lookup_field_values(THD *, COND *, TABLE_LIST *, LOOKUP_FIELD_VALUES *);
 
 /***************************************************************************
@@ -2395,7 +2360,7 @@ void Show_explain_request::call_in_target_thread()
   DBUG_ASSERT(current_thd == target_thd);
   set_current_thd(request_thd);
   if (target_thd->lex->print_explain(explain_buf, 0 /* explain flags*/,
-                                     &printed_anything))
+                                     false /*TODO: analyze? */, &printed_anything))
   {
     failed_to_produce= TRUE;
   }
@@ -4676,6 +4641,10 @@ public:
                   from frm files and storage engine are filled by the function
                   get_all_tables().
 
+  @note           This function assumes optimize_for_get_all_tables() has been
+                  run for the table and produced a "read plan" in 
+                  tables->is_table_read_plan.
+
   @param[in]      thd                      thread handler
   @param[in]      tables                   I_S table
   @param[in]      cond                     'WHERE' condition
@@ -4692,16 +4661,16 @@ int get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
   TABLE_LIST table_acl_check;
   SELECT_LEX *lsel= tables->schema_select_lex;
   ST_SCHEMA_TABLE *schema_table= tables->schema_table;
-  LOOKUP_FIELD_VALUES lookup_field_vals;
+  IS_table_read_plan *plan= tables->is_table_read_plan;
   enum enum_schema_tables schema_table_idx;
   Dynamic_array<LEX_STRING*> db_names;
-  COND *partial_cond= 0;
+  Item *partial_cond= plan->partial_cond;
   int error= 1;
   Open_tables_backup open_tables_state_backup;
 #ifndef NO_EMBEDDED_ACCESS_CHECKS
   Security_context *sctx= thd->security_ctx;
 #endif
-  uint table_open_method;
+  uint table_open_method= tables->table_open_method;
   bool can_deadlock;
   DBUG_ENTER("get_all_tables");
 
@@ -4725,9 +4694,6 @@ int get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
   thd->reset_n_backup_open_tables_state(&open_tables_state_backup);
 
   schema_table_idx= get_schema_table_idx(schema_table);
-  tables->table_open_method= table_open_method=
-    get_table_open_method(tables, schema_table, schema_table_idx);
-  DBUG_PRINT("open_method", ("%d", tables->table_open_method));
   /* 
     this branch processes SHOW FIELDS, SHOW INDEXES commands.
     see sql_parse.cc, prepare_schema_table() function where
@@ -4751,44 +4717,12 @@ int get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
     goto err;
   }
 
-  if (get_lookup_field_values(thd, cond, tables, &lookup_field_vals))
+  if (plan->no_rows)
   {
     error= 0;
     goto err;
   }
 
-  DBUG_PRINT("info",("db_name='%s', table_name='%s'",
-                     lookup_field_vals.db_value.str,
-                     lookup_field_vals.table_value.str));
-
-  if (!lookup_field_vals.wild_db_value && !lookup_field_vals.wild_table_value)
-  {
-    /*
-      if lookup value is empty string then
-      it's impossible table name or db name
-    */
-    if ((lookup_field_vals.db_value.str &&
-         !lookup_field_vals.db_value.str[0]) ||
-        (lookup_field_vals.table_value.str &&
-         !lookup_field_vals.table_value.str[0]))
-    {
-      error= 0;
-      goto err;
-    }
-  }
-
-  if (lookup_field_vals.db_value.length &&
-      !lookup_field_vals.wild_db_value)
-    tables->has_db_lookup_value= TRUE;
-  if (lookup_field_vals.table_value.length &&
-      !lookup_field_vals.wild_table_value)
-    tables->has_table_lookup_value= TRUE;
-
-  if (tables->has_db_lookup_value && tables->has_table_lookup_value)
-    partial_cond= 0;
-  else
-    partial_cond= make_cond_for_info_schema(cond, tables);
-
   if (lex->describe)
   {
     /* EXPLAIN SELECT */
@@ -4798,7 +4732,7 @@ int get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
 
   bzero((char*) &table_acl_check, sizeof(table_acl_check));
 
-  if (make_db_list(thd, &db_names, &lookup_field_vals))
+  if (make_db_list(thd, &db_names, &plan->lookup_field_vals))
     goto err;
   for (size_t i=0; i < db_names.elements(); i++)
   {
@@ -4814,7 +4748,7 @@ int get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
     {
       Dynamic_array<LEX_STRING*> table_names;
       int res= make_table_name_list(thd, &table_names, lex,
-                                    &lookup_field_vals, db_name);
+                                    &plan->lookup_field_vals, db_name);
       if (res == 2)   /* Not fatal error, continue */
         continue;
       if (res)
@@ -4852,8 +4786,8 @@ int get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
             already created by make_table_name_list() function).
           */
           if (!table_open_method && schema_table_idx == SCH_TABLES &&
-              (!lookup_field_vals.table_value.length ||
+              (!plan->lookup_field_vals.table_value.length ||
-               lookup_field_vals.wild_table_value))
+               plan->lookup_field_vals.wild_table_value))
           {
             table->field[0]->store(STRING_WITH_LEN("def"), system_charset_info);
             if (schema_table_store_record(thd, table))
@@ -8030,6 +7964,137 @@ int make_schema_select(THD *thd, SELECT_LEX *sel,
 }
 
 
+/*
+  Optimize reading from an I_S table.
+
+  @detail
+    This function prepares a plan for populating an I_S table with 
+    get_all_tables().
+
+    The plan is in IS_table_read_plan structure, it is saved in
+    tables->is_table_read_plan.
+
+  @return
+    false - Ok
+    true  - Out Of Memory
+    
+*/
+
+static bool optimize_for_get_all_tables(THD *thd, TABLE_LIST *tables, COND *cond)
+{
+  SELECT_LEX *lsel= tables->schema_select_lex;
+  ST_SCHEMA_TABLE *schema_table= tables->schema_table;
+  enum enum_schema_tables schema_table_idx;
+  IS_table_read_plan *plan;
+  DBUG_ENTER("get_all_tables");
+
+  if (!(plan= new IS_table_read_plan()))
+    DBUG_RETURN(1);
+
+  tables->is_table_read_plan= plan;
+
+  schema_table_idx= get_schema_table_idx(schema_table);
+  tables->table_open_method= get_table_open_method(tables, schema_table, 
+                                                   schema_table_idx);
+  DBUG_PRINT("open_method", ("%d", tables->table_open_method));
+
+  /* 
+    this branch processes SHOW FIELDS, SHOW INDEXES commands.
+    see sql_parse.cc, prepare_schema_table() function where
+    this values are initialized
+  */
+  if (lsel && lsel->table_list.first)
+  {
+    /* These do not need to have a query plan */
+    goto end;
+  }
+
+  if (get_lookup_field_values(thd, cond, tables, &plan->lookup_field_vals))
+  {
+    plan->no_rows= true;
+    goto end;
+  }
+
+  DBUG_PRINT("info",("db_name='%s', table_name='%s'",
+                     plan->lookup_field_vals.db_value.str,
+                     plan->lookup_field_vals.table_value.str));
+
+  if (!plan->lookup_field_vals.wild_db_value && 
+      !plan->lookup_field_vals.wild_table_value)
+  {
+    /*
+      if lookup value is empty string then
+      it's impossible table name or db name
+    */
+    if ((plan->lookup_field_vals.db_value.str &&
+         !plan->lookup_field_vals.db_value.str[0]) ||
+        (plan->lookup_field_vals.table_value.str &&
+         !plan->lookup_field_vals.table_value.str[0]))
+    {
+      plan->no_rows= true;
+      goto end;
+    }
+  }
+
+  if (plan->has_db_lookup_value() && plan->has_table_lookup_value())
+    plan->partial_cond= 0;
+  else
+    plan->partial_cond= make_cond_for_info_schema(cond, tables);
+  
+end:
+  DBUG_RETURN(0);
+}
+
+
+/*
+  This is the optimizer part of get_schema_tables_result().
+*/
+
+bool optimize_schema_tables_reads(JOIN *join)
+{
+  THD *thd= join->thd;
+  bool result= 0;
+  DBUG_ENTER("optimize_schema_tables_reads");
+
+  for (JOIN_TAB *tab= first_linear_tab(join, WITH_CONST_TABLES); 
+       tab; 
+       tab= next_linear_tab(join, tab, WITH_BUSH_ROOTS))
+  {
+    if (!tab->table || !tab->table->pos_in_table_list)
+      continue;
+
+    TABLE_LIST *table_list= tab->table->pos_in_table_list;
+    if (table_list->schema_table && thd->fill_information_schema_tables())
+    {
+      /* A value of 0 indicates a dummy implementation */
+      if (table_list->schema_table->fill_table == 0)
+        continue;
+
+      /* skip I_S optimizations specific to get_all_tables */
+      if (table_list->schema_table->fill_table != get_all_tables)
+        continue;
+
+      Item *cond= tab->select_cond;
+      if (tab->cache_select && tab->cache_select->cond)
+      {
+        /*
+          If join buffering is used, we should use the condition that is
+          attached to the join cache. Cache condition has a part of WHERE that
+          can be checked when we're populating this table.
+          join_tab->select_cond is of no interest, because it only has
+          conditions that depend on both this table and previous tables in the
+          join order.
+        */
+        cond= tab->cache_select->cond;
+      }
+
+      optimize_for_get_all_tables(thd, table_list, cond);
+    }
+  }
+  DBUG_RETURN(result);
+}
+
+
 /*
   Fill temporary schema tables before SELECT
 
@@ -8038,6 +8103,10 @@ int make_schema_select(THD *thd, SELECT_LEX *sel,
     join  join which use schema tables
     executed_place place where I_S table processed
 
+  SEE ALSO
+    The optimization part is done by get_schema_tables_result(). This function
+    is run on query execution.
+
   RETURN
     FALSE success
     TRUE  error
@@ -8058,7 +8127,7 @@ bool get_schema_tables_result(JOIN *join,
 
   for (JOIN_TAB *tab= first_linear_tab(join, WITH_CONST_TABLES); 
        tab; 
-       tab= next_linear_tab(join, tab, WITHOUT_BUSH_ROOTS))
+       tab= next_linear_tab(join, tab, WITH_BUSH_ROOTS))
   {
     if (!tab->table || !tab->table->pos_in_table_list)
       break;

sql/sql_show.h

@@ -150,6 +150,69 @@ public:
   void call_in_target_thread();
 };
 
+/**
+  Condition pushdown used for INFORMATION_SCHEMA / SHOW queries.
+  This structure is to implement an optimization when
+  accessing data dictionary data in the INFORMATION_SCHEMA
+  or SHOW commands.
+  When the query contain a TABLE_SCHEMA or TABLE_NAME clause,
+  narrow the search for data based on the constraints given.
+*/
+typedef struct st_lookup_field_values
+{
+  /**
+    Value of a TABLE_SCHEMA clause.
+    Note that this value length may exceed @c NAME_LEN.
+    @sa wild_db_value
+  */
+  LEX_STRING db_value;
+  /**
+    Value of a TABLE_NAME clause.
+    Note that this value length may exceed @c NAME_LEN.
+    @sa wild_table_value
+  */
+  LEX_STRING table_value;
+  /**
+    True when @c db_value is a LIKE clause,
+    false when @c db_value is an '=' clause.
+  */
+  bool wild_db_value;
+  /**
+    True when @c table_value is a LIKE clause,
+    false when @c table_value is an '=' clause.
+  */
+  bool wild_table_value;
+} LOOKUP_FIELD_VALUES;  
+
+
+/*
+  INFORMATION_SCHEMA: Execution plan for get_all_tables() call
+*/
+
+class IS_table_read_plan : public Sql_alloc
+{
+public:
+  IS_table_read_plan() : no_rows(false) {}
+
+  bool no_rows;
+
+  LOOKUP_FIELD_VALUES lookup_field_vals;
+  Item *partial_cond;
+
+  bool has_db_lookup_value()
+  {
+    return (lookup_field_vals.db_value.length &&
+           !lookup_field_vals.wild_db_value);
+  }
+  bool has_table_lookup_value()
+  {
+    return (lookup_field_vals.table_value.length &&
+            !lookup_field_vals.wild_table_value);
+  }
+};
+
+bool optimize_schema_tables_reads(JOIN *join);
+
 /* Handle the ignored database directories list for SHOW/I_S. */
 bool ignore_db_dirs_init();
 void ignore_db_dirs_free();

sql/sql_update.cc

@@ -277,6 +277,7 @@ int mysql_update(THD *thd,
   List<Item> all_fields;
   killed_state killed_status= NOT_KILLED;
   Update_plan query_plan(thd->mem_root);
+  Explain_update *explain;
   query_plan.index= MAX_KEY;
   query_plan.using_filesort= FALSE;
   DBUG_ENTER("mysql_update");
@@ -717,15 +718,16 @@ int mysql_update(THD *thd,
   if (table->file->ha_table_flags() & HA_PARTIAL_COLUMN_READ)
     table->prepare_for_position();
 
+  explain= thd->lex->explain->get_upd_del_plan();
   /*
     We can use compare_record() to optimize away updates if
     the table handler is returning all columns OR if
     if all updated columns are read
   */
   can_compare_record= records_are_comparable(table);
-
   while (!(error=info.read_record(&info)) && !thd->killed)
   {
+    explain->on_record_read();
     if (table->vfield)
       update_virtual_fields(thd, table,
                             table->triggers ? VCOL_UPDATE_ALL :
@@ -736,6 +738,7 @@ int mysql_update(THD *thd,
       if (table->file->was_semi_consistent_read())
         continue;  /* repeat the read of the same row if it still exists */
 
+      explain->on_record_after_where();
       store_record(table,record[1]);
       if (fill_record_n_invoke_before_triggers(thd, table, fields, values, 0,
                                                TRG_EVENT_UPDATE))
@@ -993,7 +996,11 @@ int mysql_update(THD *thd,
   id= thd->arg_of_last_insert_id_function ?
     thd->first_successful_insert_id_in_prev_stmt : 0;
 
-  if (error < 0)
+  if (thd->lex->analyze_stmt)
+  {
+    error= thd->lex->explain->send_explain(thd);
+  }
+  else if (error < 0)
   {
     char buff[MYSQL_ERRMSG_SIZE];
     my_snprintf(buff, sizeof(buff), ER(ER_UPDATE_INFO), (ulong) found,

sql/sql_yacc.yy

@@ -982,7 +982,7 @@ bool my_yyoverflow(short **a, YYSTYPE **b, ulong *yystacksize);
 
    This makes the code grep-able, and helps maintenance.
 */
-
+ 
 %token  ABORT_SYM                     /* INTERNAL (used in lex) */
 %token  ACCESSIBLE_SYM
 %token  ACTION                        /* SQL-2003-N */
@@ -1804,6 +1804,7 @@ bool my_yyoverflow(short **a, YYSTYPE **b, ulong *yystacksize);
 %type <dyncol_def_list> dyncall_create_list
 
 %type <NONE>
+        analyze_stmt_command
         query verb_clause create change select do drop insert replace insert2
         insert_values update delete truncate rename
         show describe load alter optimize keycache preload flush
@@ -1990,6 +1991,7 @@ verb_clause:
 statement:
           alter
         | analyze
+        | analyze_stmt_command
         | binlog_base64_event
         | call
         | change
@@ -12766,6 +12768,13 @@ describe_command:
         | DESCRIBE
         ;
 
+analyze_stmt_command:
+          ANALYZE_SYM explainable_command
+          {
+            Lex->analyze_stmt= true;
+          }
+        ;
+
 opt_extended_describe:
           /* empty */ {}
         | EXTENDED_SYM   { Lex->describe|= DESCRIBE_EXTENDED; }

sql/table.h

@@ -1499,6 +1499,7 @@ typedef struct st_schema_table
   uint i_s_requested_object;  /* the object we need to open(TABLE | VIEW) */
 } ST_SCHEMA_TABLE;
 
+class IS_table_read_plan;
 
 /*
   Types of derived tables. The ending part is a bitmap of phases that are
@@ -2044,12 +2045,23 @@ struct TABLE_LIST
   /* TRUE <=> this table is a const one and was optimized away. */
   bool optimized_away;
 
+  /* I_S: Flags to open_table (e.g. OPEN_TABLE_ONLY or OPEN_VIEW_ONLY) */
   uint i_s_requested_object;
-  bool has_db_lookup_value;
+
-  bool has_table_lookup_value;
+  /*
+    I_S: how to read the tables (SKIP_OPEN_TABLE/OPEN_FRM_ONLY/OPEN_FULL_TABLE)
+  */
   uint table_open_method;
+  /*
+    I_S: where the schema table was filled
+    (this is a hack. The code should be able to figure out whether reading
+    from I_S should be done by create_sort_index() or by JOIN::exec.)
+  */
   enum enum_schema_table_state schema_table_state;
 
+  /* Something like a "query plan" for reading INFORMATION_SCHEMA table */
+  IS_table_read_plan *is_table_read_plan;
+
   MDL_request mdl_request;
 
 #ifdef WITH_PARTITION_STORAGE_ENGINE