Re: bk commit into 5.0 tree (svoj:1.2526) BUG#29734

[Guilhem Bichot]

Hi,

On Fri, Jul 13, 2007 at 06:55:41PM +0500, Sergey Vojtovich wrote:
> ChangeSet@stripped, 2007-07-13 18:55:36+05:00, svoj@stripped +1 -0
>   BUG#29734 - thread_id=0 in binary log which leads to temporary table conflicts
>   
>   thread_id was not reset properly after mysql_change_user() command, which
>   may cause replication failure when replicating temporary tables.
>   
>   Fixed by resetting thread_id to proper value after mysql_change_user().

wrong cset comment

>   sql/sql_class.cc@stripped, 2007-07-13 18:55:34+05:00, svoj@stripped +1 -0
>     Fixed that pseudo_thread_id was set to 0 after mysql_change_user().
> 
> diff -Nrup a/sql/sql_class.cc b/sql/sql_class.cc
> --- a/sql/sql_class.cc	2007-07-06 03:43:02 +05:00
> +++ b/sql/sql_class.cc	2007-07-13 18:55:34 +05:00
> @@ -328,6 +328,7 @@ void THD::init(void)
>  					       variables.date_format);
>    variables.datetime_format= date_time_format_copy((THD*) 0,
>  						   variables.datetime_format);
> +  variables.pseudo_thread_id= thread_id;
>    pthread_mutex_unlock(&LOCK_global_system_variables);
>    server_status= SERVER_STATUS_AUTOCOMMIT;
>    if (variables.sql_mode & MODE_NO_BACKSLASH_ESCAPES)
> 

Ok to push.
Please remove "variables.pseudo_thread_id= 0" from THD::THD(), because
THD::THD() calls THD::init(), and you have now added a setting of
pseudo_thread_id in THD::init().

-- 
   __  ___     ___ ____  __
  /  |/  /_ __/ __/ __ \/ /    Mr. Guilhem Bichot <guilhem@stripped>
 / /|_/ / // /\ \/ /_/ / /__   MySQL AB, Lead Software Engineer
/_/  /_/\_, /___/\___\_\___/   Bordeaux, France
       <___/   www.mysql.com