Hi all,
please review the fix for 8042581. This failure was only observed
on Windows, so it is most possibly because the socket.close() on Windows
is processed asynchronously within the Windows API layer (TCP machine) .
Add 5 times of tries to the test if BindException is encountered on Win
Looks ok Felix.
You could put the construction of the new DatagramSocket in a
try-with-resources, so it will be closed always, even if the test fails.
-Chris
> On 29 Dec 2014, at 10:04, FELIX YANG wrote:
>
> Hi all,
> please review the fix for 8042581. This failure was only observed on
>
On 29/12/2014 15:46, Chris Hegarty wrote:
Looks ok Felix.
You could put the construction of the new DatagramSocket in a
try-with-resources, so it will be closed always, even if the test fails.
-Chris
I agree with Chris' comment. One other thing is that testing if e
instanceof BindException i
On 27 Dec 2014, at 15:53, Doychin Bondzhev wrote:
> This is a copy of my email that I sent earlier to core-libs-dev.
> ---
>
> Hi,
>
> I want to report for a problem that I see when I have more then one IP
> addresses assigned to same interface.
>
> Here is an output from ip addr com
Yes, It is only a matter of implementation in the native code. There
will be no changes in the public API.
To be honest I don't have much experience in C development and I'm not
sure the function that I'm offering is supported on all UNIX platforms
that OpenJDK targets.
I can test only on L
Thanks for looking at this, Alan and Chris. Please check the updated
webrev:
http://cr.openjdk.java.net/~xiaofeya/8042581/webrev.01/
-Felix
On 12/29/2014 11:51 PM, Alan Bateman wrote:
On 29/12/2014 15:46, Chris Hegarty wrote:
Looks ok Felix.
You could put the construction of the new Datagram
