Valgrind warning from hp_clear_dataspace/hp_free_level

Bug #783583 reported by Philip Stoev on 2011-05-16
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Philip Stoev

Bug Description

The DDL RQG workload produced the following valgrind warning:

==8706== Address 0x18c853b8 is 504 bytes inside a block of size 127,920 free'd
==8706== at 0x4A04D72: free (vg_replace_malloc.c:325)
==8706== by 0x8D360A: my_free (my_malloc.c:128)
==8706== by 0x969855: hp_free_level (hp_block.c:146)
==8706== by 0x966F62: hp_clear_dataspace (hp_dspace.c:171)
==8706== by 0x9698BF: hp_clear (hp_clear.c:33)
==8706== by 0x96B2DC: hp_free (hp_create.c:478)
==8706== by 0x96B146: heap_try_free (hp_create.c:437)
==8706== by 0x96B1B8: heap_delete_table (hp_create.c:452)
==8706== by 0x961075: ha_heap::delete_table(char const*) (
==8706== by 0x76642F: handler::ha_delete_table(char const*) (
==8706== by 0x763A26: ha_delete_table(THD*, handlerton*, char const*, char const*, char const*, bool) (
==8706== by 0x678F1C: quick_rm_table(handlerton*, char const*, char const*, unsigned int) (
==8706== by 0x6847D3: mysql_alter_table(THD*, char*, char*, st_ha_create_information*, TABLE_LIST*, Alter_info*, unsigned int, st_order*, bool) (
==8706== by 0x8ADAC2: Alter_table_statement::execute(THD*) (
==8706== by 0x5FC1F3: mysql_execute_command(THD*) (
==8706== by 0x5FED7A: mysql_parse(THD*, char*, unsigned int, Parser_state*) (

Changed in percona-projects-qa:
assignee: nobody → Philip Stoev (pstoev-askmonty)

The valgrind log looks incomplete, i.e. it only has the second half where it says what the failing address is close to, but does not have the first half, which points to the failing operation itself.

Changed in percona-projects-qa:
status: New → In Progress
Changed in percona-projects-qa:
milestone: none → 5.5.13-eb

Philip, what is the status of this?

Changed in percona-projects-qa:
status: In Progress → Incomplete
importance: Undecided → High
Philip Stoev (pstoev-askmonty) wrote :

This was only seen once and never again. You can close this bug if you want, I will reopen it if I get a test case.

Changed in percona-projects-qa:
status: Incomplete → Invalid
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers