Details
-
Bug
-
Status: Closed (View Workflow)
-
Blocker
-
Resolution: Fixed
-
10.6
Description
bb-10.6-release 02e97252 |
safe_mutex: Found wrong usage of mutex 'LOCK_thd_data' and 'wait_mutex'
|
Mutex currently locked (in reverse order):
|
wait_mutex /data/src/10.6/storage/innobase/handler/ha_innodb.cc line 5024
|
LOCK_thd_data /data/src/10.6/sql/sql_class.h line 3851
|
LOCK_thd_kill /data/src/10.6/sql/sql_class.h line 3850
|
The failure started happening after this commit in 10.6
commit e039720bf3494a35b34eb0ddc55af170a1807723
|
Author: Marko Mäkelä
|
Date: Mon Sep 11 14:51:02 2023 +0300
|
|
MDEV-32096 Parallel replication lags because innobase_kill_query() may fail to interrupt a lock wait
|
That's expected, as the mutex lock was only added there.
Since the same test doesn't trigger any other failures for me on a build before the commit, I'll consider it a regression for now. Feel free to demote if the analysis shows otherwise.
Attachments
Issue Links
- is caused by
-
MDEV-32530 Race condition in lock_wait_rpl_report()
- Closed