diff --git a/mysql-test/r/hash_join_weedout.result b/mysql-test/r/hash_join_weedout.result
new file mode 100644
index 00000000000..72d96baa9fa
--- /dev/null
+++ b/mysql-test/r/hash_join_weedout.result
@@ -0,0 +1,41 @@
+CREATE TABLE a (
+pk INTEGER,
+col_varchar VARCHAR(1),
+col_varchar_key VARCHAR(1)
+)
+partition by hash(pk) partitions 11;
+CREATE TABLE b (
+pk INTEGER NOT NULL AUTO_INCREMENT,
+col_varchar VARCHAR(1),
+col_varchar_key VARCHAR(1),
+PRIMARY KEY (pk),
+KEY varchar_key (col_varchar_key)
+)
+partition by key() partitions 11;
+INSERT INTO a VALUES (1, 'N', '0');
+INSERT INTO b VALUES (1, '8', 'r'), (2, 'v', 'C'), (3, 'b', 'p'), (4, '7', 'W');
+SELECT 1 FROM (b AS table1
+INNER JOIN a AS table2 ON table2.pk = table1.pk OR table1.col_varchar < 'D')
+WHERE (NOT EXISTS
+(SELECT 1 FROM (b AS alias3 STRAIGHT_JOIN a AS alias4
+ON alias4.col_varchar = alias3.col_varchar_key)
+WHERE alias3.pk >= table1.pk));
+1
+1
+1
+1
+INSERT INTO a VALUES (1, 'N', '0');
+SELECT 1 FROM (b AS table1
+INNER JOIN a AS table2 ON table2.pk = table1.pk OR table1.col_varchar < 'D')
+WHERE (NOT EXISTS
+(SELECT 1 FROM (b AS alias3 STRAIGHT_JOIN a AS alias4
+ON alias4.col_varchar = alias3.col_varchar_key)
+WHERE alias3.pk >= table1.pk));
+1
+1
+1
+1
+1
+1
+1
+DROP TABLE a, b;
diff --git a/mysql-test/t/hash_join_weedout.test b/mysql-test/t/hash_join_weedout.test
new file mode 100644
index 00000000000..a638f866188
--- /dev/null
+++ b/mysql-test/t/hash_join_weedout.test
@@ -0,0 +1,36 @@
+CREATE TABLE a (
+  pk INTEGER,
+  col_varchar VARCHAR(1),
+  col_varchar_key VARCHAR(1)
+)
+partition by hash(pk) partitions 11;
+
+CREATE TABLE b (
+  pk INTEGER NOT NULL AUTO_INCREMENT,
+  col_varchar VARCHAR(1),
+  col_varchar_key VARCHAR(1),
+  PRIMARY KEY (pk),
+  KEY varchar_key (col_varchar_key)
+)
+partition by key() partitions 11;
+INSERT INTO a VALUES (1, 'N', '0');
+INSERT INTO b VALUES (1, '8', 'r'), (2, 'v', 'C'), (3, 'b', 'p'), (4, '7', 'W');
+
+SELECT 1 FROM (b AS table1
+  INNER JOIN a AS table2 ON table2.pk = table1.pk OR table1.col_varchar < 'D')
+  WHERE (NOT EXISTS
+    (SELECT 1 FROM (b AS alias3 STRAIGHT_JOIN a AS alias4
+      ON alias4.col_varchar = alias3.col_varchar_key)
+      WHERE alias3.pk >= table1.pk));
+
+INSERT INTO a VALUES (1, 'N', '0');
+
+SELECT 1 FROM (b AS table1
+  INNER JOIN a AS table2 ON table2.pk = table1.pk OR table1.col_varchar < 'D')
+  WHERE (NOT EXISTS
+    (SELECT 1 FROM (b AS alias3 STRAIGHT_JOIN a AS alias4
+      ON alias4.col_varchar = alias3.col_varchar_key)
+      WHERE alias3.pk >= table1.pk));
+
+## clean up
+DROP TABLE a, b;
diff --git a/sql/hash_join_buffer.cc b/sql/hash_join_buffer.cc
index de180a20108..53c21052d44 100644
--- a/sql/hash_join_buffer.cc
+++ b/sql/hash_join_buffer.cc
@@ -235,7 +235,7 @@ static bool ShouldCopyRowId(const hash_join_buffer::Table &tbl) {
   // It is not safe to copy the row ID if we have a NULL-complemented row; the
   // value is undefined, or the buffer location can even be nullptr.
   const TABLE *table = tbl.qep_tab->table();
-  return tbl.rowid_status != NO_ROWID_NEEDED && !table->const_table &&
+  return tbl.qep_tab->rowid_status != NO_ROWID_NEEDED && !table->const_table &&
          !(table->is_nullable() && table->null_row);
 }
 
diff --git a/sql/hash_join_iterator.cc b/sql/hash_join_iterator.cc
index a3236507253..8498cd94308 100644
--- a/sql/hash_join_iterator.cc
+++ b/sql/hash_join_iterator.cc
@@ -101,12 +101,14 @@ HashJoinIterator::HashJoinIterator(
     if (it.qep_tab->rowid_status == NEED_TO_CALL_POSITION_FOR_ROWID) {
       it.qep_tab->rowid_status = ROWID_PROVIDED_BY_ITERATOR_READ_CALL;
     }
+    it.qep_tab->hj_iterator = this;
   }
 
   for (const hash_join_buffer::Table &it : m_probe_input_table.tables()) {
     if (it.qep_tab->rowid_status == NEED_TO_CALL_POSITION_FOR_ROWID) {
       it.qep_tab->rowid_status = ROWID_PROVIDED_BY_ITERATOR_READ_CALL;
     }
+    it.qep_tab->hj_iterator = this;
   }
 }
 
@@ -290,7 +292,7 @@ static bool WriteRowToChunk(
 void RequestRowId(const Prealloced_array<hash_join_buffer::Table, 4> &tables) {
   for (const hash_join_buffer::Table &it : tables) {
     TABLE *table = it.qep_tab->table();
-    if (it.rowid_status == NEED_TO_CALL_POSITION_FOR_ROWID &&
+    if (it.qep_tab->rowid_status == ROWID_PROVIDED_BY_ITERATOR_READ_CALL &&
         can_call_position(table)) {
       table->file->position(table->record[0]);
     }
diff --git a/sql/sql_executor.h b/sql/sql_executor.h
index 9ff9f1faa84..51c1f652583 100644
--- a/sql/sql_executor.h
+++ b/sql/sql_executor.h
@@ -50,6 +50,7 @@ class Field;
 class Field_longlong;
 class Filesort;
 class FollowTailIterator;
+class HashJoinIterator;
 class Item;
 class Item_sum;
 class JOIN;
@@ -795,6 +796,11 @@ class QEP_TAB : public QEP_shared_owner {
    */
   FollowTailIterator *recursive_iterator = nullptr;
 
+  /**
+    If this table is belong to a hash join iterator, store hj_iterator here.
+  */
+  HashJoinIterator *hj_iterator = nullptr;
+
   QEP_TAB(const QEP_TAB &);             // not defined
   QEP_TAB &operator=(const QEP_TAB &);  // not defined
 };
diff --git a/sql/sql_select.cc b/sql/sql_select.cc
index 01d2fdfb92b..7de1e092b44 100644
--- a/sql/sql_select.cc
+++ b/sql/sql_select.cc
@@ -1129,6 +1129,10 @@ SJ_TMP_TABLE *create_sj_tmp_table(THD *thd, JOIN *join,
     }
     qep_tab->table()->prepare_for_position();
     qep_tab->rowid_status = NEED_TO_CALL_POSITION_FOR_ROWID;
+    // adjust the rowid_status, which will used in hash join
+    if (qep_tab->hj_iterator != nullptr) {
+      qep_tab->rowid_status = ROWID_PROVIDED_BY_ITERATOR_READ_CALL;
+    }
   }
 
   SJ_TMP_TABLE *sjtbl;