MySQL Lists are EOL. Please join:

List:Commits« Previous MessageNext Message »
From:Davi Arnaut Date:November 10 2009 6:14pm
Subject:bzr commit into mysql-5.5-next-mr-runtime branch (davi:2916) Bug#32140
View as plain text  
# At a local mysql-5.5-next-mr-runtime repository of davi

 2916 Davi Arnaut	2009-11-10
      Backport of Bug#32140 to mysql-next-mr
      revno: 2618
      revision-id: sp1r-davi@stripped/endora.local-20080418131946-26951
      parent: sp1r-davi@stripped/endora.local-20080417190810-26185
      committer: davi@stripped/endora.local
      timestamp: Fri 2008-04-18 10:19:46 -0300
        Bug#32140: wrong error code caught when an SF() call is interruped with KILL query
        The problem is that killing a query which calls a stored function
        could return a wrong error (table corrupt) instead of the query
        interrupted error message.
        The solution is to not set the table corrupt error if the query
        is killed, the query interrupted error message will be set  later
        when the query is finished.
     @ sql/
        Don't set a error if the thread was killed, the query
        interrupted error will be set later.

=== modified file 'sql/'
--- a/sql/	2009-10-14 16:37:38 +0000
+++ b/sql/	2009-11-10 18:14:53 +0000
@@ -1711,6 +1711,9 @@ sp_cache_routines_and_add_tables_aux(THD
         ret= SP_OK;
+        /* Query might have been killed, don't set error. */
+        if (thd->killed)
+          break;
           Any error when loading an existing routine is either some problem
           with the mysql.proc table, or a parse error because the contents

Attachment: [text/bzr-bundle] bzr/
bzr commit into mysql-5.5-next-mr-runtime branch (davi:2916) Bug#32140Davi Arnaut10 Nov