[prev in list] [next in list] [prev in thread] [next in thread] 

List:       openjdk-2d-dev
Subject:    Re: [OpenJDK 2D-Dev] <AWT Dev> JDK 9 RFR of JDK-8032047: Fix static lint warnings in client librarie
From:       Joe Darcy <joe.darcy () oracle ! com>
Date:       2014-01-18 0:01:30
Message-ID: 52D9C45A.3050300 () oracle ! com
[Download RAW message or body]

Hi Sergey,

On 01/17/2014 03:45 PM, Sergey Bylokhov wrote:
> Fix looks fine.
> Does this option will apply to the open and closed builds? Since 
> openjdk can use some source code, which was not compiled in oraclejdk.

The javac options set in $JDK/make/Setup.gmk apply to both the open and 
closed sources associated with the jdk repository, including any 
platform-dependent code.

Therefore, when increasing the set of lint warnings, doing 
cross-platform closed builds is required too in case some of that code 
is afflicted by the warning in questions. My setup I typically use for 
development is an OpenJDK-only Linux laptop.

A build on my laptop succeeds with the latest patch for 8032047 and the 
jprt job would succeed too with the additional change in 
D3DSurfaceData.java.

I *think* that should cover all the relevant build combinations. Do you 
think another flavor of build needs to be done too?

Thanks,

-Joe

>
> On 18.01.2014 3:24, Joe Darcy wrote:
>> Hello,
>>
>> An update, good news / bad news from the jprt job; the jprt job 
>> failed, but there was only a single missed instance of a static 
>> warning in the sources to the jdk repo:
>>
>> src\windows\classes\sun\java2d\d3d\D3DSurfaceData.java:741: warning: 
>> [static] static method should be qualified by type name, 
>> D3DRenderQueue, instead of by an expression
>>         if (rq.isRenderQueueThread()) {
>>               ^
>>
>> I plan to fix it accordingly:
>>
>> @@ -738,7 +738,7 @@
>>          D3DRenderQueue rq = D3DRenderQueue.getInstance();
>>          // swapBuffers can be called from the toolkit thread by 
>> swing, we
>>          // should detect this and prevent the deadlocks
>> -        if (rq.isRenderQueueThread()) {
>> +        if (D3DRenderQueue.isRenderQueueThread()) {
>>              if (!rq.tryLock()) {
>>                  // if we could not obtain the lock, repaint the area
>>                  // that was supposed to be swapped, and no-op this swap
>>
>> Given that the builds were otherwise successful, I plan to get a fix for
>>
>>         JDK-8032048: Add static lint warning to build of jdk repository
>>
>> reviewed and then push JDK-8032047 and JDK-8032048 together in a 
>> single jprt job.
>>
>> Thanks,
>>
>> -Joe
>>
>>
>> On 01/17/2014 02:50 PM, Joe Darcy wrote:
>>> Hello,
>>>
>>> I've had a jprt job submitted to make sure that aren't any 
>>> cross-platform build issues (and to flush out any other static 
>>> warnings in the code base).
>>>
>>> If the jprt is successful, I'll push the changes proposed so far.
>>>
>>> Thanks Petr and Sergey for the reviews,
>>>
>>> -Joe
>>>
>>> On 01/17/2014 02:12 PM, Petr Pchelko wrote:
>>>> Hello, Joe.
>>>>
>>>> The final version still looks good.
>>>>
>>>> With best regards. Petr.
>>>>
>>>> 18 янв. 2014 г., в 1:47 до полудня, Sergey Bylokhov 
>>>> <Sergey.Bylokhov@oracle.com> написал(а):
>>>>
>>>>> Hello, Joe.
>>>>> This version looks good to me too.
>>>>>
>>
>
>

[prev in list] [next in list] [prev in thread] [next in thread]