RFR (7xS): 8071280: Specialize HeapRegion::oops_on_card_seq_iterate_careful() for use during concurrent refinement and updating the rset

Erik Helin erik.helin at oracle.com
Thu Jun 1 09:35:31 UTC 2017


On 05/31/2017 10:14 AM, Thomas Schatzl wrote:
> Hi all,
>
> On Tue, 2017-05-23 at 13:23 +0200, Thomas Schatzl wrote:
>> Hi all,
>>
>>   Erik Helin had a few comments regarding naming etc. that this new
>> webrev incorporates:
>>
>> Webrevs:
>> http://cr.openjdk.java.net/~tschatzl/8071280/webrev.2_to_3/ (diff)
>> http://cr.openjdk.java.net/~tschatzl/8071280/webrev.3/ (full)
>
>   Erik and me did some more investigation on this and found that
> actually the is_gc_active parameter
> for HeapRegion::is_obj_dead_with_size() is not required any more due to
> the addition of the ClassUnloadingWithConcurrentMark clause inside.
>
> I removed that and updated the webrev.
>
> Sorry for another change - Erik promised me that it's good now :)

Yep, this is good to go now. As we've discussed, there are some further 
improvements that can be done to this code, but we do that as a later patch.

Thanks a lot for cleaning up this code!
Erik

> Thanks,
>   Thomas
>



More information about the hotspot-gc-dev mailing list