diff options
author | Sergei Petrunia <sergey@mariadb.com> | 2022-12-22 15:49:10 +0300 |
---|---|---|
committer | Sergei Petrunia <sergey@mariadb.com> | 2023-01-04 16:50:12 +0300 |
commit | 0bef50e50b59ea1d85fea027f45cc51ca74e4074 (patch) | |
tree | b358f442e3e4a48a443a7cf732ab2a37bbac04d6 /sql/sql_union.cc | |
parent | 5db970fc760d4cfebdff236ba5bb382f0419b9f1 (diff) | |
download | mariadb-git-bb-10.4-mdev20501-v2.tar.gz |
MDEV-20501: Assertion `maybe_null || !null_value' failed in Item_func_round::date_opbb-10.4-mdev20501-v2
When the optimizer finds a constant (or system) table $TBL which is empty
or has no matching row, it would set table->null_row=true. This is done
even for tables with table->maybe_null==0.
Then, it would proceed to perform query optimization phases (what for?)
which will attempt to evaluate Item expressions referring to $TBL.
Eventually some Item expression will get the value of $TBL.not_null_field,
get SQL NULL and fail an assertion.
Fixed by not performing any query optimization steps after we've got
constant/empty tables with no matching rows.
Test result changes contain a lot of changes like
- ... Impossible WHERE noticed after reading const tables
+ ... no matching row in const table
as well as other changes caused by slightly-different processing of
the special case of empty constant tables.
Diffstat (limited to 'sql/sql_union.cc')
-rw-r--r-- | sql/sql_union.cc | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/sql/sql_union.cc b/sql/sql_union.cc index 0f218b73b16..286adb8b257 100644 --- a/sql/sql_union.cc +++ b/sql/sql_union.cc @@ -1740,6 +1740,7 @@ bool st_select_lex_unit::exec() if (describe) { /* + UglySubqueryReoptimization: In EXPLAIN command, constant subqueries that do not use any tables are executed two times: - 1st time is a real evaluation to get the subquery value |