<AWT Dev> [9] Review Request: 7124271 [macosx] RealSync test failure

Sergey Bylokhov Sergey.Bylokhov at oracle.com
Mon Aug 17 12:17:30 UTC 2015

Any volunteers to review the fix?

On 11.08.15 21:29, Alexander Zvegintsev wrote:
> Hello Sergey,
> the fix looks good to me.
> -- 
> Thanks,
> Alexander.
> On 10.08.2015 14:21, Sergey Bylokhov wrote:
>> Hello.
>> Please review the fix for jdk9.
>> Two bugs were fixed:
>>  - The changes in NSApplicationAWT.waitForDummyEvent() is a fix for a 
>> regression of JDK-8080504, which does not take into account that we 
>> should not call an unlock if a lock operation was failed with timeout.
>>  - Initial bug which was filed is about use of two different API to 
>> post events to the application, [NSApplication postEvent] and 
>> CGEventPost(). It was found that if we post an event via CGEventPost 
>> in robot and immediately after this we will post the second event via 
>> [NSApp postEvent] in realsync then sometimes the second event will be 
>> handled first. The opposite isn't proved, but I changed the code to 
>> use both to be safer.
>> Bug: https://bugs.openjdk.java.net/browse/JDK-7124271
>> Webrev can be found at: 
>> http://cr.openjdk.java.net/~serb/7124271/webrev.00

Best regards, Sergey.

More information about the awt-dev mailing list