[MDEV-12731] rpl.rpl_gtid_stop_start fails with Valgrind in buildbot and outside Created: 2017-05-08  Updated: 2017-09-08  Resolved: 2017-09-08

Status: Closed
Project: MariaDB Server
Component/s: Replication, Tests
Affects Version/s: 10.2
Fix Version/s: 10.2.9

Type: Bug Priority: Major
Reporter: Elena Stepanova Assignee: Sachin Setiya (Inactive)
Resolution: Fixed Votes: 0
Labels: None

Issue Links:
Relates
relates to MDEV-11621 rpl.rpl_gtid_stop_start fails sporadi... Closed
relates to MDEV-12629 Conditional jump or move depends on u... Closed
Sprint: 10.2.7-1

 Description   

http://buildbot.askmonty.org/buildbot/builders/work-amd64-valgrind/builds/10030/steps/test/logs/stdio

rpl.rpl_gtid_stop_start 'innodb,mix'     [ fail ]  Found warnings/errors in server log file!
        Test ended at 2017-05-08 20:40:06
line
==26785== Conditional jump or move depends on uninitialised value(s)
==26785==    at 0xB37740: find_stage_class(unsigned int) (pfs_instr_class.cc:1080)
==26785==    by 0xB65D85: start_stage_v1 (pfs.cc:4307)
==26785==    by 0x5D1E57: inline_mysql_set_stage(unsigned int, char const*, int) (mysql_stage.h:65)
==26785==    by 0x5E3823: THD::enter_stage(PSI_stage_info_v1 const*, char const*, char const*, unsigned int) (sql_class.h:2184)
==26785==    by 0x67D897: THD::exit_cond(PSI_stage_info_v1 const*, char const*, char const*, int) (sql_class.h:3160)
==26785==    by 0x600FBC: start_slave_thread(unsigned int, void* (*)(void*), st_mysql_mutex*, st_mysql_mutex*, st_mysql_cond*, unsigned int volatile*, unsigned long volatile*, Master_info*) (slave.cc:966)
==26785==    by 0x60131D: start_slave_threads(THD*, bool, bool, Master_info*, char const*, char const*, int) (slave.cc:1053)
==26785==    by 0x5FFDCC: init_slave() (slave.cc:531)
==26785==    by 0x5DBA8D: mysqld_main(int, char**) (mysqld.cc:5968)
==26785==    by 0x5D02EF: main (main.cc:25)
==26957== Conditional jump or move depends on uninitialised value(s)
==26957==    at 0xB37740: find_stage_class(unsigned int) (pfs_instr_class.cc:1080)
==26957==    by 0xB65D85: start_stage_v1 (pfs.cc:4307)
==26957==    by 0x5D1E57: inline_mysql_set_stage(unsigned int, char const*, int) (mysql_stage.h:65)
==26957==    by 0x5E3823: THD::enter_stage(PSI_stage_info_v1 const*, char const*, char const*, unsigned int) (sql_class.h:2184)
==26957==    by 0x67D897: THD::exit_cond(PSI_stage_info_v1 const*, char const*, char const*, int) (sql_class.h:3160)
==26957==    by 0x600FBC: start_slave_thread(unsigned int, void* (*)(void*), st_mysql_mutex*, st_mysql_mutex*, st_mysql_cond*, unsigned int volatile*, unsigned long volatile*, Master_info*) (slave.cc:966)
==26957==    by 0x60131D: start_slave_threads(THD*, bool, bool, Master_info*, char const*, char const*, int) (slave.cc:1053)
==26957==    by 0x5FFDCC: init_slave() (slave.cc:531)
==26957==    by 0x5DBA8D: mysqld_main(int, char**) (mysqld.cc:5968)
==26957==    by 0x5D02EF: main (main.cc:25)
^ Found warnings in /data/bld/10.2-valgrind/mysql-test/var/log/mysqld.2.err



 Comments   
Comment by Andrei Elkin [ 2017-07-06 ]

Monty, hi.

Could you please review a single-liner patch.
The issue is rather well-known in mysql as well. I refer to in the commit message.

Thanks!

Andrei

Comment by Andrei Elkin [ 2017-07-06 ]

Hi.

Could you please review a single-liner.
The issue is rather well-known in mysql. Yet I found that as late as I had arrived at solution..

There is at least on duplicate which I grabbed from you MDEV-12629.

The patch must be present in bb-10.2-andrei.

Thanks!

Andrei

Comment by Andrei Elkin [ 2017-07-06 ]

Hi.

Could you please review the task.
A single-liner patch is in bb-10.2-andrei.

Cheers,

Andrei

Comment by Sachin Setiya (Inactive) [ 2017-08-25 ]

Okay to push.

Generated at Thu Feb 08 08:00:01 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.