RFR of JDK-7195382: TEST_BUG: java/rmi/activation/CommandEnvironment/SetChildEnv.java can fail

[Hamlin Li]

Hi Roger,

Thank you for revewing.

On 2016/12/9 23:36, Roger Riggs wrote:
> Hi Hamlin,
>
> The patch looks fine but I would wrap the long @run lines.
I was thinking it's easily for reviewer to compare different @run's with 
long lines, but modified as you suggested.
>
> Looking "-Djava.compiler=NONE" made me wonder why that was important 
> to the test.
I had same thought, it's removed.
The code is pushed after modified as you suggested.

Thank you
-Hamlin
>
> Roger
>
>
> On 12/8/2016 10:07 PM, Hamlin Li wrote:
>> Is some one available to review the patch?
>>
>> Thank you
>> -Hamlin
>>
>> On 2016/12/8 11:07, Hamlin Li wrote:
>>> Would you please review the below patch?
>>>
>>> bug: https://bugs.openjdk.java.net/browse/JDK-7195382
>>> webrev: http://cr.openjdk.java.net/~mli/7195382/webrev.00/
>>>
>>> what's changed?
>>> * Use createRMIDOnEphemeralPort to avoid "port in use" exception.
>>> * Because createRMIDOnEphemeralPort will choose different ports for 
>>> rmid for multiple runs, it will fail subsequent ActivationGroup 
>>> except of first run, so have to run 4 test cases in different JVM by 
>>> putting them in rather in separate "@run main/othervm".
>>> * Put rmid.cleanup() in finally block to avoid orphan process bug.
>>> * delete dead code.
>>>
>>> Thank you
>>> -Hamlin
>>
>