diff options
| author | Evgeny Potemkin <evgeny.potemkin@oracle.com> | 2012-06-28 16:53:45 +0400 |
|---|---|---|
| committer | Evgeny Potemkin <evgeny.potemkin@oracle.com> | 2012-06-28 16:53:45 +0400 |
| commit | 0f7888f7cae3edac5e168aea44ba39b99aa29026 (patch) | |
| tree | d915ad66e31deb0d20812509f7ea3e51889aa20e /sql/sql_update.cc | |
| parent | b128b1d42444c736b72af72e04131c782a03d769 (diff) | |
| download | mariadb-git-0f7888f7cae3edac5e168aea44ba39b99aa29026.tar.gz | |
Bug#14248833: UPDATE ON INNODB TABLE ENTERS RECURSION
Introduction of cost based decision on filesort vs index for UPDATE
statements changed detection of the fact that the index used to scan the
table is being updated. The new design missed the case of index merge
when there is no single index to check. That was worked until a recent
change in InnoDB after which it went into infinite recursion if update of
the used index wasn't properly detected.
The fix consists of 'used key being updated' detection code from 5.1.
Diffstat (limited to 'sql/sql_update.cc')
| -rw-r--r-- | sql/sql_update.cc | 9 |
1 files changed, 9 insertions, 0 deletions
diff --git a/sql/sql_update.cc b/sql/sql_update.cc index 754170e4c55..c4a95edcfc2 100644 --- a/sql/sql_update.cc +++ b/sql/sql_update.cc @@ -450,6 +450,15 @@ int mysql_update(THD *thd, { // Check if we are modifying a key that we are used to search with: used_key_is_modified= is_key_used(table, used_index, table->write_set); } + else if (select && select->quick) + { + /* + select->quick != NULL and used_index == MAX_KEY happens for index + merge and should be handled in a different way. + */ + used_key_is_modified= (!select->quick->unique_key_range() && + select->quick->is_keys_used(table->write_set)); + } #ifdef WITH_PARTITION_STORAGE_ENGINE if (used_key_is_modified || order || |