Fwd: RFR: 8158946 - btree009 fails with assert(s > 0) failed: Bad size calculated

Thomas Schatzl thomas.schatzl at oracle.com
Mon Jul 4 12:18:48 UTC 2016


Hi Derek,

On Thu, 2016-06-30 at 17:15 -0400, Derek White wrote:
> New webrev goes back to using java_lang_Class::set_oop_size(new_cls,
> size);
> 
> http://cr.openjdk.java.net/~drwhite/8158946/webrev.04/
> 
> jprt run in progress. Built w/ and w/o PCH.

- the assert in java_lang_Class::oop_size() could be a bit improved,
maybe use the same message as for oopDesc::size_given_class() for the
same check, maybe the same in java_lang_class::set_oop_size()

- I am not sure if the comment above CollectedHeap::class_allocate()
makes any sense at that location. This is far from where it is actually
set and seems to be only a random comment.

I do not need a re-review for these fix-ups.

Looks good.

Thanks,
  Thomas




More information about the hotspot-runtime-dev mailing list