<Swing Dev> [14] RFR JDK-8234149:Several regression tests do not dispose Frame at end

Prasanta Sadhukhan prasanta.sadhukhan at oracle.com
Fri Nov 22 09:30:52 UTC 2019


ping? any more feedback on this?

On 18-Nov-19 4:18 PM, Prasanta Sadhukhan wrote:
> I have removed the manual test updation. Rest are passing and mach5 is 
> green (except many nothing to run tests which are either problemlisted 
> or platform-specific)
>
> http://cr.openjdk.java.net/~psadhukhan/8234149/webrev.1/
>
> Regards
>
> Prasanta
>
> On 18-Nov-19 9:53 AM, Sergey Bylokhov wrote:
>> On 11/14/19 11:02 pm, Prasanta Sadhukhan wrote:
>>> Hi Sergey,
>>>
>>> I can remove the dispose from manual tests, but are you questioning 
>>> the basic need of these dispose in "all" tests?
>>
>> Only about manual tests.
>>
>>>
>>> On 15-Nov-19 12:17 PM, Sergey Bylokhov wrote:
>>>> Hi, Prasanta.
>>>>
>>>> In a few files you have updated the manual tests which I think 
>>>> changed the behavior of the tests, for example:
>>>> test/jdk/javax/swing/JSlider/6524424/bug6524424.java
>>>>
>>>> Not sure that such code is useful:
>>>>   59         frame.setVisible(true);
>>>>   60         } finally {
>>>>   61             if (frame != null) SwingUtilities.invokeAndWait(() 
>>>> -> frame.dispose());
>>>>   62         }
>>>>
>>>> On 11/14/19 12:56 am, Prasanta Sadhukhan wrote:
>>>>> Hi All,
>>>>>
>>>>> Please review a fix for an issue seen in several tests where the 
>>>>> JFrame created is not disposed of at all code paths.
>>>>>
>>>>> This sometimes leaves some artifacts thereby affecting execution 
>>>>> of subsequent tests, causing those to fail.
>>>>>
>>>>> Proposed fix is to make sure the JFrame is disposed in try-finally 
>>>>> block to ensure frame is disposed irrespective of execution 
>>>>> outcome of the test.
>>>>>
>>>>> This list is not exhaustive as there are still some tests it is 
>>>>> not covered here and
>>>>>
>>>>> I have not taken care of indentation to ensure only the 
>>>>> change/modification is highlighted and not all indented lines.
>>>>>
>>>>> Mach5 status link is there in JBS.
>>>>>
>>>>> Bug: https://bugs.openjdk.java.net/browse/JDK-8234149
>>>>>
>>>>> webrev: http://cr.openjdk.java.net/~psadhukhan/8234149/webrev.0/
>>>>>
>>>>> Regards
>>>>>
>>>>> Prasanta
>>>>>
>>>>
>>>>
>>
>>


More information about the swing-dev mailing list