Thanks Jorgen.
I wonder, should I give mysql_derived_cleanup() a more precise name,
emphasizing that it might not clean up in certain situations?
Jorgen Loland a écrit, Le 06.06.2011 12:51:
> Hi Guilhem,
>
> I would prefer a simpler test case, but the code looks good. Approved.
>
> On 06/01/2011 05:19 PM, Guilhem Bichot wrote:
>> #At
>> file:///home/mysql_src/bzrrepos_new/mysql-next-mr-opt-backporting-wl4800/
>> based on revid:guilhem.bichot@stripped
>>
>> 3322 Guilhem Bichot 2011-06-01
>> fix for BUG#12612201 - SEGFAULT IN
>> SUBSELECT_UNIQUESUBQUERY_ENGINE::PRINT WITH OPTIMIZER TRACE
>> @ sql/sql_derived.cc
>> In mysql_derived_cleanup(), don't clean up SELECT_LEX_UNIT or
>> JOIN, if
>> EXPLAIN or optimizer trace. Cleanup happens later then.
>> This was already the case for some mysql_derived_cleanup() calls
>> in SELECT, this is now extended to UPDATE (because optimizer
>> trace
>> works for UPDATE too). It will also be useful for EXPLAIN
>> UPDATE.
