Request for review: 6312706: Map entrySet iterators should return different entries on each call to next()

Mike Duigou mike.duigou at
Tue Mar 29 00:42:38 UTC 2011

On Mar 28 2011, at 14:01 , Mike Duigou wrote:
> On Mar 26 2011, at 01:10 , Neil Richards wrote:
>> Hi Mike - thanks for posting up the webrev, and for your review.
>> On Fri, 2011-03-25 at 19:33 -0700, Mike Duigou wrote:
>>> I have run the jtreg regression suite against your webrev. I had to
>>> make one change to IdentityHashMap to satisfy the Collection/MOAT
>>> test. The failure condition involved calling EntrySet.remove() when
>>> there was no current entry due to a prior remove() or because next()
>>> had never been called. 
>> With the assignment line suitably corrected, I believe there is no need
>> for the added check.
>> Please find below a corrected changeset 'hg export' with the relevant
>> correction made.
> I will apply this patch and retest.

The updated changeset passes all the jtreg regression tests. I have updated the webrev:

I don't have any remaining concerns with this patch. My residual anxiety regarding possible performance slowdown is fading. :-)

If there are no objections I will commit this changeset on Wednesday March 30th, 2011.


More information about the core-libs-dev mailing list