RFR(XS) 8242789: sun/tools/jhsdb/HeapDumpTestWithActiveProcess.java fails with 'JShellToolProvider' missing from stdout/stderr

[Alex Menkov]

+1

--alex

On 04/20/2020 12:03, serguei.spitsyn at oracle.com wrote:
> Hi Chris,
> 
> LGTM
> 
> Thanks,
> Serguei
> 
> 
> On 4/20/20 10:46, Chris Plummer wrote:
>> Ping. This is a very simple change.
>>
>> thanks,
>>
>> Chris
>>
>> On 4/17/20 10:30 AM, Chris Plummer wrote:
>>> Hello,
>>>
>>> Please review the following:
>>>
>>> https://bugs.openjdk.java.net/browse/JDK-8242789
>>> http://cr.openjdk.java.net/~cjplummer/8242789/webrev.00
>>>
>>> JShellHeapDumpTest.java has two variants, one that does a short 2 
>>> second sleep after launching the jshell process (the main 
>>> JShellHeapDumpTest.java test does this) and the other that does no 
>>> sleep (HeapDumpTestWithActiveProcess.java does this by invoking 
>>> JShellHeapDumpTest.java with the "nosleep" argument).
>>>
>>> The reason for the 2 second sleep is to get the jshell process into a 
>>> steady state so JDK-8231634 [1] doesn't turn up when using SA on the 
>>> jshell process. I added the sleep instead of problem listing 
>>> JShellHeapDumpTest.java since it is a useful test even with the sleep 
>>> in place. HeapDumpTestWithActiveProcess.java was added so we still 
>>> had a test to reproduce JDK-8231634 [1], and was problem listed 
>>> immediately. However, another side affect of not sleeping is 
>>> sometimes SA requests the thread dump of the jshell process before 
>>> jshell enters its main thread. Thus the test can't find the 
>>> "JShellToolProvider" symbol in the thread dump. The fix is to simply 
>>> not require the symbol to be present when in "nosleep" mode.
>>>
>>> thanks,
>>>
>>> Chris
>>>
>>> [1] https://bugs.openjdk.java.net/browse/JDK-8231634
>>>
>>
>>
>