Review request for 8022208: Intermittent test failures in java/lang/Thread/ThreadStateTest.java

Mandy Chung mandy.chung at oracle.com
Thu Oct 31 13:41:59 PDT 2013


On 10/31/2013 12:34 PM, Alan Bateman wrote:
> On 31/10/2013 19:11, Mandy Chung wrote:
>> Updated webrev that has a new 
>> test/lib/testlibrary/ThreadStateController.java and also change to 
>> use AtomicInteger:
>> http://cr.openjdk.java.net/~mchung/jdk8/webrevs/8022208/webrev.01/
>>
>> Mandy
> Is ThreadStateController general enough for the unname package in 
> /lib/testlibrary? Just wondering if it should be moved down into a 
> package (I like using testlibrary, just wonder how it will scale if we 
> make over-use of it).
>

I was also wondering how testlibrary is currently structured.  While 
ThreadStateController can be for general use, I would think that this 
class would be more specific for these tests and other ThreadMXBean 
tests to use. Having a second thought, it may be better to move it under 
jdk/test/java/lang/Thread/testlibrary?  Any opinion?

> Also does the refactoring mean that methods can be removed from 
> ThreadMXBean/Utils.java? I realize this is expanding your scope a bit 
> and maybe changing SynchronizationStatistics (which appears to be the 
> user other of Utils).
>

I looked at them and I decide to leave it as a separate task.

ThreadMXBean/Utils and ThreadExecutionSynchronizer are currently used by 
a few other ThreadMXBean tests that are intermittent test failures (I am 
responsible for them :).  Jarsolav has fixed several of them and he has 
also filed a bug to remove ThreadExecutionSynchronizer and use 
java.util.concurrent where appropriate.
    https://bugs.openjdk.java.net/browse/JDK-8024326

I may spare some time to help clean them up but no promise :)

Mandy
Mandy

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/serviceability-dev/attachments/20131031/81bafcc5/attachment.html 


More information about the serviceability-dev mailing list