Commit 0992be92 authored by Vicențiu Ciorbaru's avatar Vicențiu Ciorbaru

MDEV-13068 Crash in Item::split_sum_func2 with INSERT SELECT using window functions

When running setup fields during the final step of insert using select
the final setup_fields does not have any sum functions. Our current
condition for calling split_sum_func however would attempt to use an empty
NULL sum_func_list, if the item contained a window function.

The solution is to not perform another split_sum_func for the item
containing a window function if we do not actually have a sum_func_list.
parent c73fa2d7
CREATE TABLE t1 (c1 INT, c2 VARCHAR(30));
PREPARE populate_table FROM "INSERT into t1 values (1, 'manual_insert_1'),
(4, 'manual_insert_2')";
INSERT INTO t1 SELECT row_number() over(), "should_have_0" FROM t1;
INSERT INTO t1 SELECT 1 + row_number() over(), "should_have_2" FROM t1;
EXECUTE populate_table;
INSERT INTO t1 SELECT 10 + row_number() over(), "should repeat 4 times [11-14]" FROM t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
1 manual_insert_1
4 manual_insert_2
11 should repeat 4 times [11-14]
12 should repeat 4 times [11-14]
13 should repeat 4 times [11-14]
14 should repeat 4 times [11-14]
0 should_have_0
2 should_have_2
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 10 + (dense_rank() over(order by c1)), "dense_rank_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
11 dense_rank_insert
12 dense_rank_insert
1 manual_insert_1
4 manual_insert_2
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 100 + (rank() over(order by c1)), "rank_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
1 manual_insert_1
4 manual_insert_2
101 rank_insert
102 rank_insert
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 100 + (ntile(10) over(order by c1)), "ntile_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
1 manual_insert_1
4 manual_insert_2
101 ntile_insert
102 ntile_insert
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 1000 + (percent_rank() over(order by c1)), "percent_rank_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
1 manual_insert_1
4 manual_insert_2
1000 percent_rank_insert
1001 percent_rank_insert
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 1000 + (count(*) over(order by c1)), "count_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
1001 count_insert
1002 count_insert
1 manual_insert_1
4 manual_insert_2
DELETE FROM t1;
EXECUTE populate_table;
#
# Test how avg interacts when the results need to be rounded.
#
SELECT 1000 + (avg(c1) over(order by c1)) as avg_expr, c1, "This will be inserted into t1" from t1;
avg_expr c1 This will be inserted into t1
1001.0000 1 This will be inserted into t1
1002.5000 4 This will be inserted into t1
INSERT INTO t1
SELECT 1000 + (avg(c1) over(order by c1)), "avg_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
c1 c2
1001 avg_insert
1003 avg_insert
1 manual_insert_1
4 manual_insert_2
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 1000 + (sum(c1) over(order by c1)), "sum_insert" from t1;
SELECT c1, c2
FROM t1
ORDER BY c2, c1;
c1 c2
1 manual_insert_1
4 manual_insert_2
1001 sum_insert
1005 sum_insert
DROP table t1;
CREATE TABLE t1 (c1 INT, c2 VARCHAR(30));
PREPARE populate_table FROM "INSERT into t1 values (1, 'manual_insert_1'),
(4, 'manual_insert_2')";
INSERT INTO t1 SELECT row_number() over(), "should_have_0" FROM t1;
INSERT INTO t1 SELECT 1 + row_number() over(), "should_have_2" FROM t1;
EXECUTE populate_table;
INSERT INTO t1 SELECT 10 + row_number() over(), "should repeat 4 times [11-14]" FROM t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 10 + (dense_rank() over(order by c1)), "dense_rank_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 100 + (rank() over(order by c1)), "rank_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 100 + (ntile(10) over(order by c1)), "ntile_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 1000 + (percent_rank() over(order by c1)), "percent_rank_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 1000 + (count(*) over(order by c1)), "count_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
--echo #
--echo # Test how avg interacts when the results need to be rounded.
--echo #
SELECT 1000 + (avg(c1) over(order by c1)) as avg_expr, c1, "This will be inserted into t1" from t1;
INSERT INTO t1
SELECT 1000 + (avg(c1) over(order by c1)), "avg_insert" from t1;
SELECT c1, c2 FROM t1 ORDER BY c2, c1;
DELETE FROM t1;
EXECUTE populate_table;
INSERT INTO t1
SELECT 1000 + (sum(c1) over(order by c1)), "sum_insert" from t1;
SELECT c1, c2
FROM t1
ORDER BY c2, c1;
DROP table t1;
...@@ -7068,10 +7068,13 @@ bool setup_fields(THD *thd, Ref_ptr_array ref_pointer_array, ...@@ -7068,10 +7068,13 @@ bool setup_fields(THD *thd, Ref_ptr_array ref_pointer_array,
split_sum_func() must be called for Window Function items, see split_sum_func() must be called for Window Function items, see
Item_window_func::split_sum_func. Item_window_func::split_sum_func.
*/ */
if ((item->with_sum_func && item->type() != Item::SUM_FUNC_ITEM && if (sum_func_list &&
sum_func_list) || item->with_window_func) ((item->with_sum_func && item->type() != Item::SUM_FUNC_ITEM) ||
item->with_window_func))
{
item->split_sum_func(thd, ref_pointer_array, *sum_func_list, item->split_sum_func(thd, ref_pointer_array, *sum_func_list,
SPLIT_SUM_SELECT); SPLIT_SUM_SELECT);
}
thd->lex->current_select->select_list_tables|= item->used_tables(); thd->lex->current_select->select_list_tables|= item->used_tables();
thd->lex->used_tables|= item->used_tables(); thd->lex->used_tables|= item->used_tables();
thd->lex->current_select->cur_pos_in_select_list++; thd->lex->current_select->cur_pos_in_select_list++;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment