On 07/04/11 09:19, Tatiana Azundris Nurnberg wrote:
> #At file:///Users/tnurnberg/forest/11902767/51-11902767/ based on
> revid:dao-gang.qu@stripped
>
> 3537 Tatiana Azundris Nurnberg 2011-04-07
> Bug#11902767/Bug#60580: Statement improperly replicated crashes slave SQL
> thread
Hi Tatiana.
Here are my review comments..
It seems the patch is for MySQL 5.1, but the triage team has
tagged it SR55MRU. Re-triage, or just skip 5.1 push ?
> === modified file 'sql/sql_load.cc'
Update the file copyright
> --- a/sql/sql_load.cc 2010-11-30 23:32:51 +0000
> +++ b/sql/sql_load.cc 2011-04-07 07:19:34 +0000
> @@ -684,8 +684,7 @@ static bool write_execute_load_query_log
> pfields.append("`");
> pfields.append(item->name);
> pfields.append("`");
> - pfields.append("=");
> - val->print(&pfields, QT_ORDINARY);
> + pfields.append(val->name);
> }
> }
>
>
> === modified file 'sql/sql_yacc.yy'
Update the file copyright
> --- a/sql/sql_yacc.yy 2010-10-13 05:28:58 +0000
> +++ b/sql/sql_yacc.yy 2011-04-07 07:19:34 +0000
> @@ -10828,7 +10828,23 @@ field_or_var:
>
> opt_load_data_set_spec:
> /* empty */ {}
> - | SET insert_update_list {}
> + | SET load_data_set_list {}
> + ;
> +
> +load_data_set_list:
> + load_data_set_list ',' load_data_set_elem
> + | load_data_set_elem
> + ;
> +
> +load_data_set_elem:
> + simple_ident_nospvar equal remember_name expr_or_default remember_end
> + {
> + LEX *lex= Lex;
> + if (lex->update_list.push_back($1) ||
> + lex->value_list.push_back($4))
> + MYSQL_YYABORT;
> + $4->set_name($3, (uint) ($5 - $3), YYTHD->charset());
> + }
I'm actually writing a worklog that tries to get rid of
remember_name/remember_end, and let the items print themselves
in a cleaner fashion.. Anyway - this is the way to go today.. :)
So - check the version issue, update the copyright headers, and then
you're ok to push :)
--Magne
