Request for approval: 7014860: Socket.getInputStream().available() not clear for shutdown input

[Sean Chou]

I was thinking it should be a behavior related change.
It is ok to me not to back port.

Thanks Alan.


On Wed, Oct 24, 2012 at 5:41 PM, Alan Bateman <Alan.Bateman at oracle.com> wrote:
> On 24/10/2012 10:27, Sean Chou wrote:
>>
>> Hello,
>>
>>      This is request for approval to push the fix of 7014860 to jdk7u,
>> please take a look.
>>
>> Link to the bug:
>> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7014860
>>
>> Changeset in jdk8: http://hg.openjdk.java.net/jdk8/tl/jdk/rev/a5a28b040714
>>
>> The webrev for jdk7u is the same with changeset of jdk8 without any
>> change.
>> Building on windows and linux passes; jtreg test contained passes.
>>
>> Review threads:
>> http://mail.openjdk.java.net/pipermail/net-dev/2011-January/002607.html
>> http://mail.openjdk.java.net/pipermail/net-dev/2011-February/002704.html
>> http://mail.openjdk.java.net/pipermail/net-dev/2011-April/003076.html
>> http://mail.openjdk.java.net/pipermail/net-dev/2011-July/003274.html
>> http://mail.openjdk.java.net/pipermail/net-dev/2011-August/003383.html
>>
> For the record I would prefer that this change not be backported to jdk7u.
> The reason is that is involves a specification change and also a subtle
> change in behavior on both Linux and Windows that seems inappropriate to do
> in an update release.
>
> -Alan.



-- 
Best Regards,
Sean Chou