<AWT Dev> Review request for CR-7179050 : [macosx] Make LWAWT be able to run on AppKit thread

Sergey Bylokhov Sergey.Bylokhov at oracle.com
Wed Jan 16 06:34:53 PST 2013


16.01.2013 16:43, Artem Ananiev wrote:
>
> On 1/15/2013 7:07 PM, Sergey Bylokhov wrote:
>> Hi, Petr.
>> Small comments from the quick view:
>>
>> ThreadUtilities performOnMainThreadWaiting should not be called from the
>> main thread with YES. I guess it should throw the same error as before
>> the fix.
>
> I don't have a strong opinion on that, but the new behavior is 
> consistent with how +performOnMainThreadWaiting:block: is implemented.
Looks like some places can be simplified too:
1 Java_sun_lwawt_macosx_LWCToolkit_isApplicationActive
2 Java_sun_java2d_opengl_CGLGraphicsConfig_getCGLConfigInfo

Well, if nobody objects, the fix looks fine then.
>
> Thanks,
>
> Artem
>
>> 15.01.2013 18:54, Petr Pchelko wrote:
>>> Hello, AWT team.
>>>
>>> Please review a fix for the issue: [macosx] Make LWAWT be able to run
>>> on AppKit thread
>>> http://bugs.sun.com/view_bug.do?bug_id=7179050
>>>
>>> The fix is available at:
>>> http://cr.openjdk.java.net/~art/pchelko/7179050/
>>> <http://cr.openjdk.java.net/%7Eart/pchelko/7179050/>
>>>
>>> A lot of simple refactoring is done here:
>>> 1. The JNFRunLoop calls are changed to ThreadUtilities calls which do
>>> the thread check before calling JNFRunloop.
>>> 2. AWT_ASSERT_NOT_APPKIT_THREAD calls are deleted where it is safe. I
>>> did not touch printing code and Drag and Drop code as in these parts
>>> the situation is much more complicated, and it would be better to
>>> investigate them separately.
>>> 3. AWT_ASSERT_APPKIT_THREAD calls immediately following the
>>> performOnMainThread are removed because it is obvious that we are
>>> running on Appkit thread.
>>> 4. AWT_ASSERT_ANY_THREAD - this was a very surprising assertion which
>>> does nothing. Probably it was originally designed to mark methods
>>> which could be called from any thread, however now it occurs in the
>>> code only occasionally, so it only confusing. It is removed
>>> 5. The CPlatformWindow method nativeSetNSWindowSecurityWarnongPosition
>>> was private, never called and all it did was throwing an Unsupported
>>> operation exception. It is deleted.
>>> 6. A lot of code in ThreadUtilities.m was used only in CocoaComponent
>>> compatibility mode. We do not support this mode any more, so the code
>>> could be deleted. It is safe, because all the removed code was used
>>> only if the sCocoaComponentCompatibility flag was set to YES. However
>>> it is set to NO in the beginning of the file and it is an only place
>>> where this flag is used in openjdk.
>>>
>>> With best regards. Petr.
>>>
>>>
>>
>>


-- 
Best regards, Sergey.




More information about the awt-dev mailing list