<AWT Dev> [14] RFR JDK-8234385:[TESTBUG] java/awt/EventQueue/6980209/bug6980209.java fails in linux nightly

[Prasanta Sadhukhan]

On 21-Nov-19 4:34 AM, Sergey Bylokhov wrote:
> On 11/20/19 1:41 am, Prasanta Sadhukhan wrote:
>> Hi Sergey,
>>
>> I have ran for 50 iteration per platform with my change and mach5 is 
>> green (link is in JBS).
>
> And it will be red w/o the fix?
>
Yes, at least 1 of the iteration will fail. mach5 link in JBS.
>>
>> Regards
>>
>> Prasanta
>>
>> On 19-Nov-19 11:38 PM, Sergey Bylokhov wrote:
>>> Hi, Prasanta.
>>>
>>> I have checked the logs for this test on different platforms and it 
>>> seems we have 3 issues here:
>>>  - RealSync does not work properly and produce next exception:
>>>     sun.awt.SunToolkit$OperationTimedOut: 20000
>>>     at 
>>> java.desktop/sun.awt.X11.XToolkit.syncNativeQueue(XToolkit.java:2645)
>>>     at java.desktop/sun.awt.SunToolkit.realSync(SunToolkit.java:1470)
>>>     at java.desktop/sun.awt.SunToolkit.realSync(SunToolkit.java:1402)
>>>     at java.desktop/java.awt.Robot.waitForIdle(Robot.java:675)
>>>     at bug6980209.main(bug6980209.java:72)
>>>     at 
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native 
>>> Method)
>>>     at 
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
>>>     at 
>>> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
>>>     at java.base/java.lang.reflect.Method.invoke(Method.java:564)
>>>     at 
>>> com.sun.javatest.regtest.agent.MainWrapper$MainThread.run(MainWrapper.java:127)
>>>     at java.base/java.lang.Thread.run(Thread.java:830)
>>>
>>>  - Exception reported in JDK-8198615, "It seems the secondary loop 
>>> will never end"
>>>  - Timeout in the realSync(), is it possible that we need to 
>>> increase the default timeout, or
>>>    we have a bug in realSync()?
>>>
>>> BTW to check the intermittent test you need to run it more than 3 
>>> times(I guess 50 pre platform is fine).
>>>
>>>
>>> On 11/19/19 1:42 am, Prasanta Sadhukhan wrote:
>>>> Hi All,
>>>>
>>>> Please review a testbug fix for an issue seen intermittently in 
>>>> linux nightly.
>>>>
>>>> Issue is that the test times out after the test finishes without 
>>>> actually propagating the test results to jtreg harness. It seems 
>>>> like the JFrame which is created is called with EXIT_ON_CLOSE which 
>>>> will call System.exit(0) causing the vm to terminate and if the 
>>>> test is not run with othervm option, it might create some issue 
>>>> when run in a bunch, as it is done in nightly testing.
>>>>
>>>> Proposed fix is to remove the EXIT_ON_CLOSE option as we already 
>>>> have frame.dispose() call. mach5 run is green.
>>>>
>>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8234385
>>>>
>>>> diff -r 9db62a092725 
>>>> test/jdk/java/awt/EventQueue/6980209/bug6980209.java
>>>> --- a/test/jdk/java/awt/EventQueue/6980209/bug6980209.java Thu Nov 
>>>> 14 10:59:45 2019 +0530
>>>> +++ b/test/jdk/java/awt/EventQueue/6980209/bug6980209.java Tue Nov 
>>>> 19 15:06:03 2019 +0530
>>>> @@ -63,7 +63,6 @@
>>>>                   public void run() {
>>>>                       frame = new JFrame();
>>>>                       frame.setUndecorated(true);
>>>> *- frame.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);*
>>>>                       setup(frame);
>>>>                   }
>>>>               });
>>>>
>>>> Regards
>>>>
>>>> Prasanta
>>>>
>>>
>>>
>
>