RFR: Event based tracing is missing thread exit

Erik Gahlin erik.gahlin at oracle.com
Thu Jul 25 14:26:30 PDT 2013


Thanks for the review, and for providing the description and the link.

Erik

Daniel D. Daugherty skrev 7/25/13 5:45 PM:
> On 7/25/13 9:21 AM, Erik Gahlin wrote:
>> Hello,
>>
>> Could you be so kind and review this small change. The fix is 
>> targeted for jdk7u40, and the hook is needed to have a leak free 
>> event tracing implementation.
>>
>> http://cr.openjdk.java.net/~egahlin/8021353/
>
> Thumbs up!
>
> The bug link in the webrev is for the old Monaco system which was
> internal only. Looks like the bug is so new that this link doesn't
> work either:
>
> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8021353
>
> I'm using this internal link:
>
> https://jbs.oracle.com/bugs/browse/JDK-8021353
>
> Here's the description from the bug report:
>
> > Event based tracing should have a hook so resources associated
> > with a thread can be cleaned up properly when a thread exits.
> > Not having such mechanism may lead to memory leaks.
>
>
> src/share/vm/runtime/thread.cpp
>     No comments.
>
> src/share/vm/trace/traceMacros.hpp
>     No comments.
>
> Dan
>
>
>>
>> Thanks
>> Erik
>



More information about the hotspot-runtime-dev mailing list