Crash Hotspot
Y. Srinivas Ramakrishna
Y.S.Ramakrishna at Sun.COM
Mon Jan 25 16:21:01 UTC 2010
Christian Thalinger wrote:
> On Mon, 2010-01-25 at 15:17 +0100, Mark Christiaens wrote:
>> Sure, I'll run with this option and report back tomorrow or so if I'm
>> still up and running. The mail you refer to mentions a bug ID, in
>> what database can I look it up? I went to bugs.sun.com but it doesn't
>> seem to know it?
>
> You mean the review ID?
>
> Review ID: 1699960
>
> External bug reports are first reviewed and then converted into bugs
> which will show up on bugs.sun.com.
>
> -- Christian
>
Just a bit more info.
The bug in the email thread that Christian referred to and
for which the workaround of -XX:-ReduceInitialCardMarks was
provided is this one:-
http://bugs.sun.com/view_bug.do?bug_id=6896647
The fix is not in JDK 7b80 (hs17-b07) that you are using,
but should show up in the next few builds (not sure which
or when, but the bug fix is in the integration stream).
The above bug is not specific to G1, but affects several other
collectors as well (but not serial gc).
As Jon mentioned there are a couple of more bugs specific to
G1 that are being actively worked on at the moment and the fixes
for some of these should appear in the next few builds.
-- ramki
More information about the hotspot-gc-dev
mailing list