Heap allocator/deallocator mismatch in scalability metrics plugin

[Laurynas Biveinis]

In 5.6:

static
sm_thd_data_t *sm_thd_data_get(MYSQL_THD thd)
{
...
    thd_data= calloc(sizeof(sm_thd_data_t), 1);
...
    thd_data->backref= list_push(thd_list_root, thd_data);
...
}

list_push will eventually do

LIST *list_cons(void *data, LIST *list)
{
  LIST *new_charset=(LIST*) my_malloc(sizeof(LIST),MYF(MY_FAE));
...

where new_charset will be stored as thd_data->backref.

But then we also have

static
void sm_thd_data_release(MYSQL_THD thd)
{
...
    thd_list_root= list_delete(thd_list_root, thd_data->backref);
...
    free(thd_data->backref);
    free(thd_data);
}

i.e. my_malloc paired with regular frees.

and we also have

      list_free(thd_list_root, TRUE);

calls, resulting in regular calloc paired with my_free.

[Laurynas Biveinis (laurynas-biveinis)]

https://github.com/percona/percona-server/pull/509, https://github.com/percona/percona-server/pull/510

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PS-981