<AWT Dev> RFC: JDK-8231991: Mouse wheel change focus on awt/swing windows

Philip Race philip.race at oracle.com
Fri Nov 15 01:30:39 UTC 2019 

OK. The incremental fix looks reasonable to me and the testing I've done 
is promising.

We are currently getting a very large number of failures (not just the 
tests we pointed out) so

let's get this pushed ASAP.

-phil.

On 11/13/19 1:39 PM, Philip Race wrote:
> Hi,
>
> I'll also apply this and submit a job to our test system.
>
> -phil.
>
> On 11/13/19, 12:46 PM, Mario Torre wrote:
>> Hi Phil,
>>
>> Ok, I got it and I feel very stupid now.
>>
>> The issue was indeed in my original patch, two problems in fact, and I
>> really don't know how I made this mistake let alone made it through
>> all the way... apparently I can't count anymore like programmers do,
>> also I inverted the logic for the isWheel.
>>
>> Here is the patch to fix this, I'm still running the whole series of
>> tests but I'm confident this time:
>>
>> http://cr.openjdk.java.net/~neugens/8234107/webrev.00/
>>
>> I apologise again for this, it shouldn't happen.
>>
>> Cheers,
>> Mario
>>
>> On Wed, Nov 13, 2019 at 8:39 PM Mario Torre
>> <neugens.limasoftware at gmail.com>  wrote:
>>> Thanks Phil, I'll spend the whole of tomorrow on this, if I can't
>>> figure it out I'll let you know ASAP so we can unblock the build.
>>>
>>> Cheers,
>>> Mario
>>>
>>> Il giorno mer 13 nov 2019 alle ore 20:33 Phil Race
>>> <philip.race at oracle.com>  ha scritto:
>>>>> I'm not sure if in the meantime you want to backout the change?
>>>> I'm not sure either.
>>>> We are trying to get a clean sheet on the Linux headful tests - 
>>>> although
>>>> there are a lot that are problem listed.
>>>>
>>>> If we can't figure this out very soon, we'll minimally have to 
>>>> problem list these tests
>>>> and have another bug open to address the cause .. well I suppose we 
>>>> need a bug regardless
>>>> so I've filed https://bugs.openjdk.java.net/browse/JDK-8234107.
>>>> That can either be used for backout, follow on fix, or problem 
>>>> listing but
>>>> if we use it for backout or problem listing someone will need to 
>>>> file a new bug
>>>> to track the re-fix.
>>>>
>>>> -phil.
>>>>
>>>>
>>>>
>>>> On 11/13/19 11:23 AM, Mario Torre wrote:
>>>>> Ok, I can reproduce the error now.
>>>>>
>>>>> The test uses the VK_TAB to traverse the buttons, I admit this 
>>>>> code is
>>>>> quite... interesting.... I'll need to debug a bit more to see why the
>>>>> focus logic for the wheel interest this. I'm not sure if in the
>>>>> meantime you want to backout the change?
>>>>>
>>>>> Cheers,
>>>>> Mario
>>>>>
>>>>> On Wed, Nov 13, 2019 at 7:28 PM Mario Torre
>>>>> <neugens.limasoftware at gmail.com>  wrote:
>>>>>> Hi Phil,
>>>>>>
>>>>>> This is strange, as I've run the full jtreg tests in addition to the
>>>>>> mouse event ones and didn't get any failure. Nonetheless, thanks for
>>>>>> spotting and apologise for breaking the build, I'll go through those
>>>>>> ASAP and see if there's a problem with the tests or something 
>>>>>> else, I
>>>>>> fail to see how this change would affect the modal dialog, since the
>>>>>> change only deal with mouse wheel event... but well...
>>>>>>
>>>>>> Cheers,
>>>>>> Mario
>>>>>>
>>>>>> Il giorno mer 13 nov 2019 alle ore 18:43 Phil Race
>>>>>> <philip.race at oracle.com>  ha scritto:
>>>>>>> Mario,
>>>>>>>
>>>>>>> Since you pushed this 4 modal dialog automated tests are failing 
>>>>>>> - mostly -
>>>>>>> on at least Ubuntu 16.04 and 18.04.
>>>>>>>
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsSetModal6Test.java
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal5Test.java 
>>>>>>>
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsDocModal2Test.java
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal6Test.java 
>>>>>>>
>>>>>>>
>>>>>>> They fail in our nightly and I reproduced it locally,
>>>>>>> They don't all always fail but never do they all pass : eg
>>>>>>>     jtreg -a -v
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsSetModal6Test.java
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal5Test.java 
>>>>>>>
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsDocModal2Test.java
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal6Test.java 
>>>>>>>
>>>>>>> runner starting test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsDocModal2Test.java
>>>>>>> runner finished test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsDocModal2Test.java
>>>>>>> Failed. Execution failed: `main' threw exception:
>>>>>>> java.lang.RuntimeException: DummyButton on Frame did not gain 
>>>>>>> focus on
>>>>>>> clicking. A document modal dialog and its parent are visible.: 
>>>>>>> expected
>>>>>>> true, was false
>>>>>>> runner starting test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsSetModal6Test.java
>>>>>>> runner finished test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsSetModal6Test.java
>>>>>>> Failed. Execution failed: `main' threw exception:
>>>>>>> java.lang.RuntimeException: DummyButton on Window did not gain 
>>>>>>> focus on
>>>>>>> clicking. The parent of the Window is null Dialog: expected 
>>>>>>> true, was false
>>>>>>> runner starting test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal5Test.java 
>>>>>>>
>>>>>>> runner finished test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal5Test.java 
>>>>>>>
>>>>>>> Failed. Execution failed: `main' threw exception:
>>>>>>> java.lang.RuntimeException: DummyButton on Window did not gain 
>>>>>>> focus on
>>>>>>> clicking. The parent of the Window is TOOLKIT_MODAL Dialog: 
>>>>>>> expected
>>>>>>> true, was false
>>>>>>> runner starting test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal6Test.java 
>>>>>>>
>>>>>>> runner finished test:
>>>>>>> java/awt/Modal/ModalBlockingTests/BlockingWindowsToolkitModal6Test.java 
>>>>>>>
>>>>>>> Passed. Execution successful
>>>>>>> Test results: passed: 1; failed: 3
>>>>>>>
>>>>>>>
>>>>>>> When I backout your change, the tests all pass again.
>>>>>>> Prasanta also reported that backing out this change fixed them.
>>>>>>>
>>>>>>> These tests aren't what I'd call 100% stable but there is 
>>>>>>> definitely a
>>>>>>> new problem here.
>>>>>>> Can you check what you see on your end and we'll have to decide if
>>>>>>> there's a follow on
>>>>>>> fix or a back out whilst we think about it, or whether you 
>>>>>>> believe there
>>>>>>> was a latent
>>>>>>> test problem ?
>>>>>>>
>>>>>>> -phil.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On 11/12/19 8:58 AM, Mario Torre wrote:
>>>>>>>> On Mon, 2019-11-11 at 11:45 +0000, Dmitry Markov wrote:
>>>>>>>>> Hi Mario,
>>>>>>>>>
>>>>>>>>> The fix looks good to me.
>>>>>>>> Thanks, I pushed it.
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>> Mario
>>>>>> -- 
>>>>>> pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
>>>>>> Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF
>>>>>>
>>>>>> Java Champion - Blog: http://neugens.wordpress.com - Twitter: 
>>>>>> @neugens
>>>>>> Proud GNU Classpath developer: http://www.classpath.org/
>>>>>> OpenJDK: http://openjdk.java.net/projects/caciocavallo/
>>>>>>
>>>>>> Please, support open standards:
>>>>>> http://endsoftpatents.org/
>>>>>>
>>>
>>> -- 
>>> pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
>>> Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF
>>>
>>> Java Champion - Blog: http://neugens.wordpress.com - Twitter: @neugens
>>> Proud GNU Classpath developer: http://www.classpath.org/
>>> OpenJDK: http://openjdk.java.net/projects/caciocavallo/
>>>
>>> Please, support open standards:
>>> http://endsoftpatents.org/
>>>
>>

More information about the awt-dev mailing list