harmony-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mikhail Markov (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HARMONY-2697) [drlvm] Un-end()-ed deflaters leak native memory
Date Wed, 23 May 2007 09:25:16 GMT

    [ https://issues.apache.org/jira/browse/HARMONY-2697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12498169
] 

Mikhail Markov commented on HARMONY-2697:
-----------------------------------------

I've linked this isssue with HARMONY-2530 which looks very similar.
The only problem is that 2530 is reproducible on J9 as well (while this JIRA report says that
it reproducible on DRLVM only) - that's why i've linked them as "reated" not "duplicated".
Nevertheless i think that fixing 2530 will resolve this issue as well.

> [drlvm] Un-end()-ed deflaters leak native memory
> ------------------------------------------------
>
>                 Key: HARMONY-2697
>                 URL: https://issues.apache.org/jira/browse/HARMONY-2697
>             Project: Harmony
>          Issue Type: Bug
>          Components: DRLVM
>         Environment: Windows XP i686
>            Reporter: Salikh Zakirov
>            Priority: Minor
>         Attachments: DeflateInflate.java
>
>
> The attached test DeflateInflate.java quickly leaks memory and crashes on DRLVM.
> The test works okay on J9 and HotSpot.
> Note, that JRockit also leaks memory and crashes.
> Effectively, the test creates huge amounts of java objects, which keep pointers to blocks
allocated
> in native heap. The native heap is exhausted before java heap is filled, so no garbage
collection
> occurs. Should the java collection occur, it would be able to detect unreachable finalizable
objects,
> run finalizers, which in turn will free native memory.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message