[
https://issues.apache.org/jira/browse/DERBY-4967?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12983673#action_12983673
]
Knut Anders Hatlen commented on DERBY-4967:
-------------------------------------------
Thanks, Dag. This way to detect the whether interruptible I/O is enabled looks
fine to me.
Is RAF.close() interruptible, by the way? I see that the interrupt flag isn't
cleared until after close() has been called. Maybe it's safer to postpone the
closing until after the interrupt flag has been cleared?
And a tiny nit: If you move "new RandomAccessFile(...)" out of the try clause
and up to the declaration, the file can be closed unconditionally in the
finally clause (can never be null).
> Handle interrupt received while waiting for database lock
> ---------------------------------------------------------
>
> Key: DERBY-4967
> URL: https://issues.apache.org/jira/browse/DERBY-4967
> Project: Derby
> Issue Type: Improvement
> Components: Store
> Reporter: Dag H. Wanvik
> Assignee: Dag H. Wanvik
> Fix For: 10.8.0.0
>
> Attachments: derby-4967-locking-1.diff, derby-4967-locking-1.stat,
> derby-4967-locking-2.diff, derby-4967-locking-2.stat,
> derby-4967-locking-3.diff, derby-4967-locking-3.stat
>
>
> Subtask of DERBY-4741: this issue tracks the changes needed to handle
> interrupt received while a thread is waiting for a database lock.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
