Comment 8 for bug 1264164

Actually, it's more then the above as reducer tries 6 times per mysqld, so it's more in the area of 1/200-400.

But, here is finally a testcase which works reasonably consistently.

[roel@localhost 598892]$ ./reducer10.sh sql.sql
2014-09-03 23:42:40 [Init] Workdir: /dev/shm/1409751760 <--- note workdir
[...]
2014-09-03 23:42:56 [Init] Looking for this string: 'm_can_overwrite_status' in mysqld error log output
[...wait for issue to reproduce... if it fails, restart reducer after rm -Rf /dev/shm/* ...]
2014-09-03 23:45:23 [] [Stage V] [MULTI] All verification subreducer threads have finished/terminated
2014-09-03 23:45:23 [*] [Stage V] [MULTI] Threads which reproduced the issue: #16
2014-09-03 23:45:23 [*] [Stage V] [MULTI] Only 1 out of 30 threads reproduced the issue: this issue is sporadic
Got it!

As soon as you see this "Got it!" you can see the occurence:

[roel@localhost 598892]$ cd /dev/shm/1409751760 <---- workdir from above
[roel@localhost 1409751760]$ grep 'm_can_overwrite_status' */*/error.log.out
mysqld: /mnt/workspace/percona-server-5.6-binaries-debug-yassl/label_exp/centos6-64/percona-server-5.6.20-68.0/sql/sql_error.cc:521: void Diagnostics_area::set_error_status(uint, const char*, const char*, const Sql_condition*): Assertion `! is_set() || m_can_overwrite_status' failed.

To kill all, press ctrl-c on reducer.sh then execute
ps -ef | grep mysqld | grep -v grep | awk '{print $2}' | xargs kill -9;

--core is passed in reducer10.sh so core file etc. are present in that directory found above. Uploading one set, as well as testcase files. Note that path for mysqld is set in reducer10.sh