RFR: 8139424: SIGSEGV, Problematic frame: # V [libjvm.so+0xd0c0cc] void InstanceKlass::oop_oop_iterate_oop_maps_specialized<true, oopDesc*, MarkAndPushClosure>

[Kim Barrett]

On Nov 12, 2015, at 3:47 AM, Stefan Johansson <stefan.johansson at oracle.com> wrote:
> 
> 
> On 2015-11-12 01:23, Kim Barrett wrote:
>> On Nov 11, 2015, at 10:41 AM, Stefan Johansson <stefan.johansson at oracle.com> wrote:
>>> Hi,
>>> 
>>> Please review this fix for:
>>> https://bugs.openjdk.java.net/browse/JDK-8139424
>>> 
>>> Webrev:
>>> http://cr.openjdk.java.net/~sjohanss/8139424/hotspot.00/
>> Good find!
>> 
>> The added guarantee in RegisterHumongousWithInCSetFastTestClosure::doHeapRegion looks like leftover debugging code.
> Thanks for looking at this Kim,
> 
> It is from the debugging, but it was left on purpose. This guarantee will help us catch the case where we iterated the old version of the SparsePRT and not the expanded one, because the iterator uses SparsePRT->_cur to iterate while occupied uses SparsePRT->_next. I could update the message to say:
> "Remembered set hash maps out of sync, cur: " SIZE_FORMAT " entries, next: " SIZE_FORMAT " entries"
> 
> I think this is more informative and I prefer to leave the guarantee in. Updated webrev:
> http://cr.openjdk.java.net/~sjohanss/8139424/hotspot.01/

It seems like this check might be more appropriately placed in the constructor for
HeapRegionRemSetIterator, for the benefit of anyone iterating over a remset.
Although that might make HeapRegionRemSet’s usage a problem.