[PATCH] bug fix for 6959636 testcase failing on windows javax/management/loading/LibraryLoader/LibraryLoaderTest.java

[Eric Wang]

Hi David,

Good catch, I have removed the try-catch block, please help to review, I 
can confirm the fix is just to run in othervm.
http://dl.dropbox.com/u/90659131/fixes/6959636/webrev/index.html

Thanks,
Eric
On 2012/9/3 13:47, David Holmes wrote:
> Hi Eric,
>
> On 3/09/2012 3:42 PM, Eric Wang wrote:
>> Hi David,
>>
>> Thanks for your comments, I have updated the fix based on your
>> suggestion, Can you please help to review again.
>> http://dl.dropbox.com/u/90659131/fixes/6959636/webrev/index.html
>
> You don't need the try/catch here:
>
>  137                 try {
>  138                     ObjectName mbean =
>  139                         ((ObjectInstance) res[j]).getObjectName();
>  140                     result = server.getAttribute(mbean, "Random");
>  141                     System.out.println("MBean #" + j + " = " + 
> mbean);
>  142                     System.out.println("Random number = " + result);
>  143                 } catch (ReflectionException e) {
>  144                     throw e;
>  145                 }
>
> And can you please confirm that the actual fix here is just to run in 
> othervm?
>
> Thanks,
> David
>
>> Thanks,
>> Eric
>> On 2012/9/3 11:45, David Holmes wrote:
>>> Hi Eric,
>>>
>>> On 3/09/2012 1:22 PM, Eric Wang wrote:
>>>> Hi All,
>>>>
>>>> I am an engineer from SQE team who is helping to cleanup ProblemList.
>>>> Here is a fix below for bug 6959636
>>>> <http://monaco.us.oracle.com/detail.jsf?cr=6959636>, Can you please 
>>>> help
>>>> to review?
>>>> http://dl.dropbox.com/u/90659131/fixes/6959636/webrev/index.html
>>>>
>>>> The test failed when it is executed by the command "jtreg -agentvm
>>>> -retain:fail,error", root cause is "-retain:fail,error" option 
>>>> tries to
>>>> remove temp files created by test but they are still opened by the
>>>> agentvm/samevm.
>>>
>>> So the fix for that is to run in othervm?
>>>
>>> And the code changes are actually nothing to do with the original
>>> problem? You are just simplifying the error logic by having the
>>> exceptions propagate?
>>>
>>> Here:
>>>
>>> 130 System.err.println("Failed to load the MBean #" + j +
>>> 131 ". The shown Throwable was caught.");
>>>
>>> and here:
>>>
>>> 145 } catch (ReflectionException e) {
>>> 146 System.err.println("A ReflectionException, wrapping " +
>>> 147 "the shown exception, occured when" +
>>>
>>> The printed messages no longer makes sense because there is no "shown
>>> exception" anymore. I'd suggest just getting rid of the println's and
>>> let the exceptions propagate directly - no need for any try/catch now
>>> that main throws Throwable.
>>>
>>> David
>>>
>>>> Thanks,
>>>> Eric
>>