[
https://issues.apache.org/jira/browse/POOL-136?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mark Thomas resolved POOL-136.
------------------------------
Resolution: Fixed
Fix Version/s: 1.5
I've fixed what I can. Use of factory needs to remain un-synced to prevent
deadlocks. This is just more fuel for the "need to rethink sync strategy in
2.0" argument.
> Inconsistent synchronization in GenericKeyedObjectPool
> ------------------------------------------------------
>
> Key: POOL-136
> URL: https://issues.apache.org/jira/browse/POOL-136
> Project: Commons Pool
> Issue Type: Bug
> Reporter: Sebb
> Fix For: 1.5
>
>
> There are several fields that have synchronized set/get methods, but which
> are not accessed in a synchronized manner.
> Access to these variables needs to be synch. when referenced from within the
> class as well.
> _minEvictableIdleTimeMillis
> _testWhileIdle - this could be made volatile, in which case get/set don't
> need to be synch.
> _totalIdle
> _totalInternalProcessing
> _whenExhaustedAction
> The setFactory() method synchronizes the code that updates the variable
> _factory; however references to the factory are not synchronized.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
