executor, cmd: refine HashJoinExec for specific JoinType if outer/inn…

…er is nil (pingcap#8296)

.gitignore

@@ -15,5 +15,6 @@ y.output
 profile.coverprofile
 explain_test
 cmd/explaintest/explain-test.out
+cmd/explaintest/explaintest_tidb-server
 _tools/
 *.fail.go

executor/join.go

@@ -211,8 +211,11 @@ func (e *HashJoinExec) fetchOuterChunks(ctx context.Context) {
 			}
 			return
 		}
-
 		if !hasWaitedForInner {
+			if outerResult.NumRows() == 0 {
+				e.finished.Store(true)
+				return
+			}
 			jobFinished, innerErr := e.wait4Inner()
 			if innerErr != nil {
 				e.joinResultCh <- &hashjoinWorkerResult{
@@ -241,7 +244,7 @@ func (e *HashJoinExec) wait4Inner() (finished bool, err error) {
 			return false, errors.Trace(err)
 		}
 	}
-	if e.hashTable.Len() == 0 && e.joinType == plannercore.InnerJoin {
+	if e.hashTable.Len() == 0 && (e.joinType == plannercore.InnerJoin || e.joinType == plannercore.SemiJoin) {
 		return true, nil
 	}
 	return false, nil

executor/join_test.go

@@ -921,3 +921,59 @@ func (s *testSuite) TestEmbeddedOuterJoin(c *C) {
 	tk.MustQuery("select * from (t1 left join t2 on t1.a = t2.a) left join (t2 t3 left join t2 t4 on t3.a = t4.a) on t2.b = 1").
 		Check(testkit.Rows("1 1 <nil> <nil> <nil> <nil> <nil> <nil>"))
 }
+
+func (s *testSuite) TestHashJoin(c *C) {
+	tk := testkit.NewTestKit(c, s.store)
+	tk.MustExec("use test")
+	tk.MustExec("drop table if exists t1, t2")
+	tk.MustExec("create table t1(a int, b int);")
+	tk.MustExec("create table t2(a int, b int);")
+	tk.MustExec("insert into t1 values(1,1),(2,2),(3,3),(4,4),(5,5);")
+	tk.MustQuery("select count(*) from t1").Check(testkit.Rows("5"))
+	tk.MustQuery("select count(*) from t2").Check(testkit.Rows("0"))
+	tk.MustExec("set @@tidb_max_chunk_size=1;")
+	result := tk.MustQuery("explain analyze select /*+ TIDB_HJ(t1, t2) */ * from t1 where exists (select a from t2 where t1.a = t2.a);")
+	// id	count	task	operator info	execution info
+	// HashLeftJoin_9	8000.00	root	semi join, inner:TableReader_13, equal:[eq(test.t1.a, test.t2.a)]	time:1.036712ms, loops:1, rows:0
+	// ├─TableReader_11	10000.00	root	data:TableScan_10	time:441.096µs, loops:1, rows:1
+	// │ └─TableScan_10	10000.00	cop	table:t1, range:[-inf,+inf], keep order:false, stats:pseudo
+	// └─TableReader_13	10000.00	root	data:TableScan_12	time:212.376µs, loops:1, rows:0
+	//   └─TableScan_12	10000.00	cop	table:t2, range:[-inf,+inf], keep order:false, stats:pseudo
+	row := result.Rows()
+	c.Assert(len(row), Equals, 5)
+	outerExecInfo := row[1][4].(string)
+	c.Assert(outerExecInfo[len(outerExecInfo)-1:], Equals, "1")
+	innerExecInfo := row[3][4].(string)
+	c.Assert(innerExecInfo[len(innerExecInfo)-1:], Equals, "0")
+
+	tk.MustExec("insert into t2 select * from t1;")
+	tk.MustExec("delete from t1;")
+	tk.MustQuery("select count(*) from t1").Check(testkit.Rows("0"))
+	tk.MustQuery("select count(*) from t2").Check(testkit.Rows("5"))
+	result = tk.MustQuery("explain analyze select /*+ TIDB_HJ(t1, t2) */ * from t1 where not exists (select a from t2 where t1.a = t2.a);")
+	// id	count	task	operator info	execution info
+	// HashLeftJoin_9	8000.00	root	anti semi join, inner:TableReader_13, equal:[eq(test.t1.a, test.t2.a)]	time:534.643µs, loops:1, rows:0
+	// ├─TableReader_11	10000.00	root	data:TableScan_10	time:35.042µs, loops:1, rows:0
+	// │ └─TableScan_10	10000.00	cop	table:t1, range:[-inf,+inf], keep order:false, stats:pseudo
+	// └─TableReader_13	10000.00	root	data:TableScan_12	time:0s, loops:0, rows:0
+	//   └─TableScan_12	10000.00	cop	table:t2, range:[-inf,+inf], keep order:false, stats:pseudo
+	row = result.Rows()
+	c.Assert(len(row), Equals, 5)
+	outerExecInfo = row[1][4].(string)
+	c.Assert(outerExecInfo[len(outerExecInfo)-1:], Equals, "0")
+	innerExecInfo = row[3][4].(string)
+	c.Assert(innerExecInfo[len(innerExecInfo)-1:], LessEqual, "5")
+
+	result = tk.MustQuery("explain select /*+ TIDB_HJ(t1, t2) */ * from t1 left outer join t2 on t1.a = t2.a;")
+	// id	count	task	operator info	execution info
+	// HashLeftJoin_6	12500.00	root	left outer join, inner:TableReader_10, equal:[eq(test.t1.a, test.t2.a)]	time:502.553µs, loops:1, rows:0
+	// ├─TableReader_8	10000.00	root	data:TableScan_7	time:27.302µs, loops:1, rows:0
+	// │ └─TableScan_7	10000.00	cop	table:t1, range:[-inf,+inf], keep order:false, stats:pseudo
+	// └─TableReader_10	10000.00	root	data:TableScan_9	time:0s, loops:0, rows:0
+	//   └─TableScan_9	10000.00	cop	table:t2, range:[-inf,+inf], keep order:false, stats:pseudo
+	c.Assert(len(row), Equals, 5)
+	outerExecInfo = row[1][4].(string)
+	c.Assert(outerExecInfo[len(outerExecInfo)-1:], Equals, "0")
+	innerExecInfo = row[3][4].(string)
+	c.Assert(innerExecInfo[len(innerExecInfo)-1:], LessEqual, "5")
+}