Valgrind: Conditional jump or move depends on uninitialised value(s) at hp_mask by check_one_key

Bug #1233131 reported by Roel Van de Paar on 2013-09-30
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Percona Server
High
Roel Van de Paar
5.1
Undecided
Unassigned
5.5
Undecided
Unassigned
5.6
High
Roel Van de Paar

Bug Description

==4246== Thread 64:
==4246== Conditional jump or move depends on uninitialised value(s)
==4246== at 0x9F3C9E: hp_mask (hp_hash.c:213)
==4246== by 0x9FA5AF: check_one_key (_check.c:137)
==4246== by 0x9FA9EE: heap_check_heap (_check.c:56)
==4246== by 0x9F9469: hp_close (hp_close.c:39)
==4246== by 0x9F95D8: heap_close (hp_close.c:28)
==4246== by 0x6409AC: handler::ha_close() (handler.cc:2594)
==4246== by 0x812968: closefrm(TABLE*, bool) (table.cc:2436)
==4246== by 0x72929E: intern_close_table(TABLE*) (sql_base.cc:821)
==4246== by 0x729477: close_thread_table(THD*, TABLE**) (sql_base.cc:1525)
==4246== by 0x72980A: close_open_tables(THD*) (sql_base.cc:1241)
==4246== by 0x729A17: close_thread_tables(THD*) (sql_base.cc:1467)
==4246== by 0x77B478: mysql_execute_command(THD*) (sql_parse.cc:5316)
==4246== by 0x780D87: mysql_parse(THD*, char*, unsigned int, Parser_state*) (sql_parse.cc:6569)
==4246== by 0x7818CC: dispatch_command(enum_server_command, THD*, char*, unsigned int) (sql_parse.cc:1436)
==4246== by 0x782604: do_command(THD*) (sql_parse.cc:1051)
==4246== by 0x8206B6: threadpool_process_request(THD*) (threadpool_common.cc:313)
==4246==
2013-09-29T04:59:17 [4418] Query: INSERT IGNORE INTO `table100_memory_int_autoinc` ( `c27` , `c31` , `c58` ) VALUES ( -6532864 , 'q' , 1 )

The INSERT IGNORE may have been the one that caused the issue. Remember also the UPDATE IGNORE we saw in bug 1168265.

Roel Van de Paar (roel11) wrote :

Not 56qual, likely shared with 5.5

description: updated
summary: - Conditional jump or move depends on uninitialised value(s) at hp_mask by
- check_one_key
+ Valgrind: Conditional jump or move depends on uninitialised value(s) at
+ hp_mask by check_one_key
Roel Van de Paar (roel11) wrote :

ps -ef | grep 'cmdrun_454' | grep -v grep | awk '{print $2}' | xargs sudo kill -9 > /dev/null 2>&1
rm -Rf /sde/967241/cmdrun_454
mkdir /sde/967241/cmdrun_454
cd /sde/randgen
bash -c "set -o pipefail; perl /sde/randgen/runall.pl --queries=100000000 --seed=867 --duration=300 --querytimeout=60 --short_column_names --reporter=Shutdown,Backtrace,QueryTimeout,ErrorLog,ErrorLogAlarm --mysqld=--log-output=none --mysqld=--sql_mode=ONLY_FULL_GROUP_BY --basedir=/sde/Percona-Server-5.6.13-rc60.5-452-debug-valgrind.Linux.x86_64 --valgrind --reporter=ValgrindErrors --validator=MarkErrorLog --mysqld=--utility-user-password=test --views --mysqld=--innodb_change_buffering=none --mysqld=--innodb_changed_pages=FORCE --mysqld=--innodb_flush_method=O_DIRECT --mysqld=--innodb_log_arch_expire_sec=120 --mysqld=--innodb_log_archive=1 --mysqld=--innodb_log_block_size=4096 --mysqld=--innodb_log_buffer_size=10485761 --mysqld=--innodb_log_file_size=10485761 --mysqld=--innodb_max_bitmap_file_size=9223372036854770000 --mysqld=--innodb_track_changed_pages=0 --mysqld=--innodb_use_global_flush_log_at_trx_commit=0 --mysqld=--loose-readonly-key-cache-division-limit=1 --mysqld=--query_cache_size=1048576 --mysqld=--query_cache_type=1 --mysqld=--readonly-key-cache-block-size=1 --mysqld=--readonly-loose-max-connect-errors=0 --mysqld=--skip-innodb_doublewrite --mysqld=--thread_handling=pool-of-threads --mysqld=--transaction-isolation=READ-UNCOMMITTED --mysqld=--userstat --mysqld=--innodb_purge_threads=32 --mysqld=--innodb_adaptive_hash_index_partitions=1 --mysqld=--innodb_buffer_pool_instances=5 --grammar=conf/percona_qa/5.6/5.6.yy --gendata=conf/percona_qa/5.6/5.6.zz3 --threads=15 --no-mask --mtr-build-thread=774 --mask=14814 --vardir1=/sde/967241/cmdrun_454 > /sde/967241/cmdrun454.l
og 2>&1"
tail -n1 /sde/967241/cmdrun454.log

Roel Van de Paar (roel11) wrote :

PS @ Percona-Server-5.6.13-rc60.5-452-debug-valgrind.Linux.x86_64
RQG @ Revno: 918 Revision-Id: <email address hidden>

Is it possible to reduce the testcase?

(sorry if we discussed this on IRC already)

tags: added: memory-se
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers