RFR(XXS): 8144629: runtime/thread/Fibonacci.java test should ran in othervm mode

[David Holmes]

On 4/12/2015 6:52 PM, Dmitry Dmitriev wrote:
> David,
>
> I think that this fix can be considered as workaround for jtreg bug.
> Running in othervm mode will give us exception with the reason rather
> than timeout error. Is it ok for you?

Fine as a workaround, but a jtreg bug needs to be filed I think.

Thanks,
David

> Thanks,
> Dmitry
>
> On 04.12.2015 10:51, David Holmes wrote:
>> Hi Dmitry,
>>
>> On 4/12/2015 7:31 AM, Dmitry Dmitriev wrote:
>>> Hello,
>>>
>>> Please review trivial fix for runtime/thread/Fibonacci.java which change
>>> execution mode to the "othervm" to avoid possible deadlock in jtreg
>>> agent when uncaught exception is got.
>>
>> While this may avoid the deadlock the issue is not with the test nor
>> the way it is run - this is a bug in jtreg, likely caused by multiple
>> test threads hitting uncaught exceptions at the same time.
>>
>> David
>>
>>> JBS: https://bugs.openjdk.java.net/browse/JDK-8144629
>>> webrev: http://cr.openjdk.java.net/~ddmitriev/8144629/webrev.00/
>>> <http://cr.openjdk.java.net/%7Eddmitriev/8144629/webrev.00/>
>>>
>>> Dmitry
>