On 19/10/2011 16:49, Kurchi Subhra Hazra wrote:
That is fine with me. Can I remove the test as a part of fixing 7102704?
Sounds good to me.
The problem is that we can't be sure what local address will be returned
on a dual stack system, other than one of the local addresses. Which
doesn't seem worth testing.
-Chris.
Thanks,
Kurchi
On 10/19/11 8:48 AM, Chris Hegarty wrote:
Michael, Kurchi,
java/net/DatagramSocket/ChangingAddress.java was added recently in
JDK8, see CR 7084030. Looking at it again I'm not sure if the test
makes sense, or at least we would have to loosen its restrictions so
much that it wouldn't be testing anything useful ( or the crooks of
the original change it was added for).
We have a very similar test in nio,
java/nio/channels/DatagramChannel/ChangingAddress.java, which already
covers similar functionality.
How about we just remove it?
-Chris.
On 19/10/2011 15:45, Michael McMahon wrote:
Hi,
We have a few test failures on the nightly test run.
So, I'd like to move them on to the problem list pending
investigation.
http://cr.openjdk.java.net/~michaelm/7102665/webrev.1/
Thanks,
Michael
