Hi Phil, I ran all client test using mach5. All related test passed, see https://mach5.us.oracle.com/mdash/jobs?search=id:dmarkov-jdk-client-20200828-1219-13803609 <https://mach5.us.oracle.com/mdash/jobs?search=id:dmarkov-jdk-client-20200828-1219-13803609>
Regards, Dmitry > On 28 Aug 2020, at 21:00, Philip Race <philip.r...@oracle.com> wrote: > > Looks Ok to me but please confirm that all related tests pass. > > -phil. > > On 8/28/20, 12:54 PM, Dmitry Markov wrote: >> >> Hello, >> >> Could you review the fix for JDK 16, please? >> >> bug: https://bugs.openjdk.java.net/browse/JDK-8252470 >> <https://bugs.openjdk.java.net/browse/JDK-8252470> >> webrev: http://cr.openjdk.java.net/~dmarkov/8252470/webrev.00/ >> <http://cr.openjdk.java.net/%7Edmarkov/8252470/webrev.00/> >> >> Problem description: >> After integration of JDK-8232114 if an IME API is invoked from a DND call a >> hang may take place. >> The DND operation runs on a toolkit thread. The IME functions should be >> executed on the toolkit thread, as well. So if, for example, the IME >> function is called from one of the methods of DropTargetListener the >> corresponding IME message will be blocked in the toolkit thread message >> queue because because the toolkit thread is busy executing DND operation. >> >> Fix: >> If DND operation is active then we shouldn’t post the IME messages to the >> toolkit thread queue and use SendMessage() function instead. >> >> Regards, >> Dmitry
