Details
-
Bug
-
Status: Closed (View Workflow)
-
Blocker
-
Resolution: Fixed
-
10.4(EOL)
Description
Caused by the MDEV-30014 patch, but not reproducible in 10.5 branch containing the same patch e.g. at 9f7da106c2c5dd9f48f8fcf40823a9d4fd655298
e.g. https://buildbot.mariadb.org/#/builders/551/builds/8139
[amd64-ubuntu-2204-valgrind | failed test (failure)]
|
rpl.create_or_replace_mix
|
spider/bugfix.checksum_table_with_quick_mode_3
|
spider/bugfix.cp932_column
|
spider/bugfix.delete_with_float_column_mariadb
|
spider/bugfix.delete_with_float_column_mysql
|
spider/bugfix.gbh_outer_fields_in_join
|
spider/bugfix.group_by_order_by_limit
|
spider/bugfix.mdev_19866
|
spider/bugfix.mdev_20100
|
spider/bugfix.mdev_20502
|
spider/bugfix.mdev_21884
|
spider/bugfix.mdev_24020
|
spider/bugfix.mdev_24343
|
spider/bugfix.mdev_24517
|
spider/bugfix.mdev_24523
|
spider/bugfix.mdev_26013
|
spider/bugfix.mdev_26544
|
spider/bugfix.mdev_28856
|
spider/bugfix.mdev_29163
|
spider/bugfix.mdev_31338
|
Seems to be caused by
[Revision 0381197855c58e339ab5034b871ea9c0c2d61522]
|
Author: Yuchen Pei <ycp@mariadb.com>
|
Date: 2023-11-17 Fri 11:48:50 AEDT
|
|
MDEV-30014 Spider should not second guess server when locking / unlocking
|
Attachments
Issue Links
- is caused by
-
MDEV-30014 heap-use-after-free in ha_spider::lock_tables(), highly sporadic SIGSEGV in intern_close_table
-
- Closed
-
- is duplicated by
-
MDEV-32895 Spider tests fail with ASAN errors in ha_spider::external_lock
-
- Closed
-
Hi holyfoot, ptal thanks
upstream/bb-10.4-mdev-32849 3cd27d5d418f7e70768c20e0f2067a3893b0e3ad
MDEV-32849 Spider: check if any table is actually locked when unlocking
This avoids the scenario in MDEV-32849, when the unlock happens after
the connection has been freed, say in rollback. This is done in 10.5+
after the commit a26700cca579926cddf9a48c45f13b32785746bb.
It may or may not prevent potential other scenarios where spider has
locked something, then for some reason the statement needs to be
rolled back and spider frees the connection, and then spider proceeds
to use the freed connection. But at least we fix the regression
introduced by MDEV-30014 to 10.4 and bring 10.4 closer in parity with
10.5+.