[jira] [Commented] (LUCENE-9738) Odd behavior with tests forked into the default ForkJoinPool

["Robert Muir (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/LUCENE-9738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17280513#comment-17280513 ] 

Robert Muir commented on LUCENE-9738:
-------------------------------------

[~dweiss] I hadn't noticed this, never having the use-case to get permissions in a fork join task.

Maybe the behavior relates to their intended use as sandboxed work units:

{noformat}
The efficiency of ForkJoinTasks stems from a set of restrictions (that are only partially statically enforceable) reflecting their main use as computational tasks calculating pure functions or operating on purely isolated objects ...  should ideally avoid synchronized methods or blocks, and should minimize other blocking synchronization apart from joining other tasks ... should also not perform blocking I/O, and should ideally access variables that are completely independent of those accessed by other running tasks. 
{noformat}

> Odd behavior with tests forked into the default ForkJoinPool
> ------------------------------------------------------------
>
>                 Key: LUCENE-9738
>                 URL: https://issues.apache.org/jira/browse/LUCENE-9738
>             Project: Lucene - Core
>          Issue Type: Bug
>            Reporter: Dawid Weiss
>            Assignee: Dawid Weiss
>            Priority: Minor
>         Attachments: LUCENE-9738.patch
>
>
> Just for the record, this will perhaps save somebody some search time. This test currently fails with out test policy:
> {code}
> public void testInsanity() throws ExecutionException, InterruptedException {
>     // This is fine. This property should be read-accessible to anything in Java.
>     System.getProperty("line.separator");
>    // but this fails:
>    ForkJoinTask<String> task =
>         ForkJoinPool.commonPool()
>             .submit(() -> System.getProperty("line.separator"));
>     // Wait a bit to avoid job stealing.
>     Thread.sleep(1000);
>     task.get();
> {code}
> The above fails with a cryptic exception:
> {code}
> Caused by: java.security.AccessControlException: access denied ("java.util.PropertyPermission" "line.separator" "read")
> 	at java.base/java.security.AccessControlContext.checkPermission(AccessControlContext.java:472)
> 	at java.base/java.security.AccessController.checkPermission(AccessController.java:1036)
> 	at java.base/java.lang.SecurityManager.checkPermission(SecurityManager.java:408)
> 	at java.base/java.lang.SecurityManager.checkPropertyAccess(SecurityManager.java:1152)
> 	at java.base/java.lang.System.getProperty(System.java:847)
> 	at org.apache.lucene.TestPermissionsInsanity.lambda$testInsanity$0(TestPermissionsInsanity.java:38)
> 	at java.base/java.util.concurrent.ForkJoinTask$AdaptedCallable.exec(ForkJoinTask.java:1453)
> 	at java.base/java.util.concurrent.ForkJoinTask.doExec(ForkJoinTask.java:290)
> 	at java.base/java.util.concurrent.ForkJoinPool$WorkQueue.topLevelExec(ForkJoinPool.java:1016)
> 	at java.base/java.util.concurrent.ForkJoinPool.scan(ForkJoinPool.java:1665)
> 	at java.base/java.util.concurrent.ForkJoinPool.runWorker(ForkJoinPool.java:1598)
> 	at java.base/java.util.concurrent.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:177)
> {code}
> What's odd here is that we only have java.base classes and our own test class on the stack. Seems like by default forkjoin executors have some kind of protection domain that disallows everything... Escaping through AccessController works but it's still surprising that it's not a built-in thing.
> Or maybe I misunderstand something, don't know...
> CC [~rmuir] - you may find this interesting.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org