Percona XtraDB Cluster - HA scalable solution for MySQL

InnoDB table lock handling problem

Reported by Seppo Jaakola on 2013-12-04
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
MySQL patches by Codership
Status tracked in 5.6
5.5
Undecided
Seppo Jaakola
5.6
Undecided
Seppo Jaakola
Percona XtraDB Cluster
Status tracked in Trunk
5.6
Undecided
Unassigned
Trunk
Undecided
Unassigned

Bug Description

Table level lock conflict resolving releases wrong lock in innodblock0lock.cc:lock_table_create()
This may be due to bad merge in some earlier InnoDB upgrade.

Issue was caught by Jan when troubleshooting: https://mariadb.atlassian.net/browse/MDEV-4223

Changed in codership-mysql:
assignee: nobody → Seppo Jaakola (seppo-jaakola)

I think the correct MDEV link here is https://mariadb.atlassian.net/browse/MDEV-4233 - relating to (lock->trx)->wait_lock == lock

Seppo Jaakola (seppo-jaakola) wrote :

Fixes pushed in:

wsrep-5.5-23: http://bazaar.launchpad.net/~codership/codership-mysql/5.5-23/revision/3943
wsrep-5.5: http://bazaar.launchpad.net/~codership/codership-mysql/wsrep-5.5/revision/3931 (+3932)
wsrep-5.6: http://bazaar.launchpad.net/~codership/codership-mysql/5.6/revision/4025

An additional fix was pushed to wsrep-5.6 in revision: http://bazaar.launchpad.net/~codership/codership-mysql/5.6/revision/4026 to grab victim trx's mutex before calling kill_victim() on table lock conflict

Regression test in MDEV-4233 passed after these fixes

To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers