3598 Sunny Bains 2011-04-21
Fix a debug assertion. The rule is wrong, the lock sys mutex is irrelevant
during rollback segment header creation.
rb://561, Approved by Marko (while reviewing 561).
3597 Sunny Bains 2011-04-21
Fix a debug assertion triggered in Michael's testing. This should not have any
impact on production code. The assertion itself was wrong. A table has locks
from several transactions and transactions have locks on several tables. The
check was not taking this into consideration. The fix is to count table locks
that are on the table that is being validated and use that in the assertion.
=== modified file 'storage/innobase/sync/sync0sync.c'
--- a/storage/innobase/sync/sync0sync.c revid:sunny.bains@stripped
+++ b/storage/innobase/sync/sync0sync.c revid:sunny.bains@stripped
@@ -1321,8 +1321,7 @@ sync_thread_add_level(
- ut_a(sync_thread_levels_contain(array, SYNC_LOCK_SYS)
- && sync_thread_levels_contain(array, SYNC_FSP_PAGE));
+ ut_a(sync_thread_levels_contain(array, SYNC_FSP_PAGE));
Attachment: [text/bzr-bundle] email@example.com
|• bzr push into mysql-trunk-innodb branch (Sunny.Bains:3597 to 3598) ||Sunny Bains||21 Apr|