mysqld innodb_information_schema segfault when using pool of threads

[Valentine Gostev]

OS: CentOS 5 x86_64

[root@gval-cent5-64-n403 mysql-55-eb-blobs]# bzr version-info
 revision-id: <email address hidden>
 date: 2011-05-30 13:48:10 +0400
 build-date: 2011-05-30 18:07:25 +0400
 revno: 3480
 branch-nick: mysql-55-eb-blobs
 [root@gval-cent5-64-n403 mysql-55-eb-blobs]# bzr pull
 Using saved parent location: http://bazaar.launchpad.net/~percona-dev/percona-server/mysql-55-eb-blobs/
 No revisions to pull.
 [root@gval-cent5-64-n403 mysql-test]# ./mtr innodb_information_schema --mysqld=--thread-handling=pool-of-threads --mysqld=--thread_pool_size=8
 Logging: ./mtr innodb_information_schema --mysqld=--thread-handling=pool-of-threads --mysqld=--thread_pool_size=8
 MySQL Version 5.5.12
 Checking supported features...
  - skipping ndbcluster
  - skipping SSL, mysqld not compiled with SSL
  - binaries are debug compiled
 Collecting tests...
 vardir: /root/mysql-55-eb-blobs/mysql-test/var
 Checking leftover processes...
 Removing old var directory...
 Creating var directory '/root/mysql-55-eb-blobs/mysql-test/var'...
 Installing system database...
 Using server port 46132

==============================================================================

TEST RESULT TIME (ms) or COMMENT
 --------------------------------------------------------------------------

worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 13000..13009
 worker[1] mysql-test-run: WARNING: running this script as _root_ will cause some tests to be skipped
 innodb.innodb_information_schema [ fail ]
         Test ended at 2011-05-30 18:05:34

CURRENT_TEST: innodb.innodb_information_schema
 mysqltest: At line 169: query 'DROP TABLE t_min, t_max, ```t'\"_str`' failed: 2013: Lost connection to MySQL server during query

The result from queries just before the failure was:
 < snip >
 X RECORD `test`.```t'\"_str` `PRIMARY` 3 '2', 'abc', '"abc', 'abc"', 'a"bc', 'a"bc"', '"abc""'
 X RECORD `test`.```t'\"_str` `PRIMARY` 3 '2', 'abc', '"abc', 'abc"', 'a"bc', 'a"bc"', '"abc""'
 X RECORD `test`.```t'\"_str` `PRIMARY` 4 '3', 'abc', '\\abc', 'abc\\', 'a\\bc', 'a\\bc\\', '\\abc\\\\'
 X RECORD `test`.```t'\"_str` `PRIMARY` 4 '3', 'abc', '\\abc', 'abc\\', 'a\\bc', 'a\\bc\\', '\\abc\\\\'
 X RECORD `test`.```t'\"_str` `PRIMARY` 5 '4', 'abc', '\0abc', 'abc\0', 'a\0bc', 'a\0bc\0', 'a\0bc\0\0'
 X RECORD `test`.```t'\"_str` `PRIMARY` 5 '4', 'abc', '\0abc', 'abc\0', 'a\0bc', 'a\0bc\0', 'a\0bc\0\0'
 X RECORD `test`.`t_min` `PRIMARY` 2 -128, 0, -32768, 0, -8388608, 0, -2147483648, 0, -9223372036854775808, 0
 X RECORD `test`.`t_min` `PRIMARY` 2 -128, 0, -32768, 0, -8388608, 0, -2147483648, 0, -9223372036854775808, 0
 X RECORD `test`.`t_max` `PRIMARY` 2 127, 255, 32767, 65535, 8388607, 16777215, 2147483647, 4294967295, 9223372036854775807, 18446744073709551615
 X RECORD `test`.`t_max` `PRIMARY` 2 127, 255, 32767, 65535, 8388607, 16777215, 2147483647, 4294967295, 9223372036854775807, 18446744073709551615
 X RECORD `test`.```t'\"_str` `PRIMARY` 1 supremum pseudo-record
 X RECORD `test`.```t'\"_str` `PRIMARY` 1 supremum pseudo-record
 lock_table COUNT(*)
 `test`.`t_max` 2
 `test`.`t_min` 2
 `test`.```t'\"_str` 10
 lock_table COUNT(*)
 "test"."t_max" 2
 "test"."t_min" 2
 "test"."`t'\""_str" 10

More results from queries before failure can be found in /root/mysql-55-eb-blobs/mysql-test/var/log/innodb_information_schema.log

Server [mysqld.1 - pid: 12374, winpid: 12374, exit: 256] failed during test run
 Server log from this test:
 ----------SERVER LOG START-----------
 110530 17:05:32 InnoDB: !!!!!!!! UNIV_DEBUG switched on !!!!!!!!!
 110530 17:05:32 InnoDB: The InnoDB memory heap is disabled
 110530 17:05:32 InnoDB: Mutexes and rw_locks use GCC atomic builtins
 110530 17:05:32 InnoDB: Compressed tables use zlib 1.2.3
 110530 17:05:32 InnoDB: Using Linux native AIO
 110530 17:05:32 InnoDB: Initializing buffer pool, size = 8.0M
 110530 17:05:32 InnoDB: Completed initialization of buffer pool
 110530 17:05:32 InnoDB: highest supported file format is Barracuda.
 110530 17:05:32 InnoDB: Waiting for the background threads to start
 110530 17:05:33 InnoDB: 1.1.6 started; log sequence number 1595675
 110530 17:05:33 [Note] Event Scheduler: Loaded 0 events
 110530 17:05:33 [Note] /root/mysql-55-eb-blobs/sql/mysqld: ready for connections.
 Version: '5.5.12-debug-log' socket: '/root/mysql-55-eb-blobs/mysql-test/var/tmp/mysqld.1.sock' port: 13000 Source distribution
 110530 17:05:34 InnoDB: Assertion failure in thread 1193482560 in file ha_innodb.cc line 1553
 InnoDB: Failing assertion: ((thd) == _current_thd())
 InnoDB: We intentionally generate a memory trap.
 InnoDB: Submit a detailed bug report to http://bugs.mysql.com.
 InnoDB: If you get repeated assertion failures or crashes, even
 InnoDB: immediately after the mysqld startup, there may be
 InnoDB: corruption in the InnoDB tablespace. Please refer to
 InnoDB: http://dev.mysql.com/doc/refman/5.5/en/forcing-innodb-recovery.html
 InnoDB: about forcing recovery.
 110530 17:05:34 - mysqld got signal 6 ;
 This could be because you hit a bug. It is also possible that this binary
 or one of the libraries it was linked against is corrupt, improperly built,
 or misconfigured. This error can also be caused by malfunctioning hardware.
 We will try our best to scrape up some info that will hopefully help diagnose
 the problem, but since we have already crashed, something is definitely wrong
 and this may fail.

key_buffer_size=1048576
 read_buffer_size=131072
 max_used_connections=10
 max_threads=9
 thread_count=10
 connection_count=10
 It is possible that mysqld could use up to
 key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 6289 K
 bytes of memory
 Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x0
 Attempting backtrace. You can use the following information to find out
 where mysqld died. If you see no messages after this, something went
 terribly wrong...
 stack_bottom = (nil) thread_stack 0x40000
 /root/mysql-55-eb-blobs/sql/mysqld(my_print_stacktrace+0x32)[0x8a064a]
 /root/mysql-55-eb-blobs/sql/mysqld(handle_segfault+0x2c8)[0x52ae72]
 /lib64/libpthread.so.0[0x7f6780857b10]
 /lib64/libc.so.6(gsignal+0x35)[0x7f677f54b265]
 /lib64/libc.so.6(abort+0x110)[0x7f677f54cd10]
 /root/mysql-55-eb-blobs/sql/mysqld[0x8d4098]
 /root/mysql-55-eb-blobs/sql/mysqld[0x8d7f63]
 /root/mysql-55-eb-blobs/sql/mysqld(_Z17ha_rollback_transP3THDb+0x13b)[0x722b7b]
 /root/mysql-55-eb-blobs/sql/mysqld(_Z14trans_rollbackP3THD+0x72)[0x6a64aa]
 /root/mysql-55-eb-blobs/sql/mysqld(_ZN3THD7cleanupEv+0x71)[0x595131]
 /root/mysql-55-eb-blobs/sql/mysqld(_Z10unlink_thdP3THD+0x4f)[0x52b353]
 /root/mysql-55-eb-blobs/sql/mysqld[0x695c3c]
 /root/mysql-55-eb-blobs/sql/mysqld(libevent_thread_proc+0x26d)[0x696389]
 /lib64/libpthread.so.0[0x7f678084f73d]
 /lib64/libc.so.6(clone+0x6d)[0x7f677f5eef6d]
 The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contains
 information that should help you find out what is causing the crash.
 Writing a core file
 ----------SERVER LOG END-------------

[Valentine Gostev (longbow)]

valgrind log

[Valentine Gostev (longbow)]

valgrind log

[Valentine Gostev (longbow)]

core dump

[Valentine Gostev (longbow)]

Files as a separate bug since condition differs from bug 781083