Thanks, I was just about to ask you about that!
Since you changed the bug to patch approved, does it mean that I do not
need a second reviewer?
I will remove the the tree lines (since it does not help on MyISAM
anyway, I added it so that a storage engine that does not check for
SQLCOM_TRUNCATE in delete_all_rows, would work if it has
reset_auto_increment implemented).
I have already filed a bug for truncating a partitioned MyISAM table
(Bug#35111)
Regards
Mattias
> diff -Nrup a/sql/ha_partition.cc b/sql/ha_partition.cc
> > --- a/sql/ha_partition.cc 2007-12-20 19:16:51 +01:00
> > +++ b/sql/ha_partition.cc 2008-03-05 23:56:03 +01:00
> > @@ -2953,13 +2980,25 @@ int ha_partition::delete_all_rows()
> > {
...
> > do
> > {
> > if ((error= (*file)->ha_delete_all_rows()))
> > DBUG_RETURN(error);
> > + /* must reset the auto_increment for some engines (eg MyISAM) */
> > + if (thd->lex->sql_command == SQLCOM_TRUNCATE)
> > + (*file)->ha_reset_auto_increment(0);
No, let's call it MyISAM bug that should be fixed in MyISAM, not in
partitioning code. Remove this workaround, please.
Sergei Golubchik wrote:
> Hi!
>
> On Mar 07, Sergei Golubchik wrote:
>> Hi!
>>
>> On Mar 05, mattiasj@stripped wrote:
>>> ChangeSet@stripped, 2008-03-05 23:56:06+01:00, mattiasj@witty. +3 -0
>>> Bug#33479: auto_increment failures in partitioning
>> thanks.
>> ok to push
>> with out change.
>
> grr, that's a very dangerous typo - completely changes the meaning.
>
> I wanted to say "with ONE change"
> (see comment in the review)
>
> Regards / Mit vielen Grüssen,
> Sergei
>
--
Mattias Jonsson, Software Engineer
MySQL AB, www.mysql.com
Are you MySQL certified? www.mysql.com/certification
