Review request: test/java/util/logging/ParentLoggersTest.java fails in samevm mode

Daniel D. Daugherty daniel.daugherty at oracle.com
Thu Jun 17 00:52:37 UTC 2010


OK by me. Forgot to say "thumbs up" before.

Dan


On 6/16/2010 6:41 PM, Mandy Chung wrote:
> Daniel D. Daugherty wrote:
>> Mandy,
>>
>> What happens when the test is run twice in samevm mode? 
>
> The test only needs to support running once in samevm mode.  Are you 
> thinking of some kind of stress testing?  In any case, I don't think 
> we need to fix the test to support running more than once in the same vm.
>
> Mandy
>
>>  
>> I suspect that the expected Logger names will be in the
>> initialLoggerNames list at the beginning of the second run
>> and thus won't be added to returnedLoggerNames on line 104.
>>
>> There isn't a way to delete a Logger so if you want this
>> test to work when run twice, you'll need to filter the
>> expected Loggers out of the initial list. So instead of
>> the defaultLogger filter on lines 48-54, you'll need to
>> filter out the expected Loggers instead.
>>
>> Dan
>>
>>
>>
>> On 6/15/2010 11:22 AM, Mandy Chung wrote:
>>> Fixed 6961408: test/java/util/logging/ParentLoggersTest.java fails 
>>> in samevm mode
>>>
>>> Webrev at:
>>>   http://cr.openjdk.java.net/~mchung/6961408/webrev.00/
>>>
>>> This fix updates the test to check the list of loggers before and 
>>> after the test run and also takes out all logging tests from the 
>>> ProblemList.txt.
>>>
>>> Thanks
>>> Mandy
>



More information about the core-libs-dev mailing list