On 21-Nov-19 4:34 AM, Sergey Bylokhov wrote:
On 11/20/19 1:41 am, Prasanta Sadhukhan wrote:
Hi Sergey,
I have ran for 50 iteration per platform with my change and mach5 is
green (link is in JBS).
And it will be red w/o the fix?
Yes, at least 1 of the iteration will fail. mach5 link in JBS.
Regards
Prasanta
On 19-Nov-19 11:38 PM, Sergey Bylokhov wrote:
Hi, Prasanta.
I have checked the logs for this test on different platforms and it
seems we have 3 issues here:
- RealSync does not work properly and produce next exception:
sun.awt.SunToolkit$OperationTimedOut: 20000
at
java.desktop/sun.awt.X11.XToolkit.syncNativeQueue(XToolkit.java:2645)
at java.desktop/sun.awt.SunToolkit.realSync(SunToolkit.java:1470)
at java.desktop/sun.awt.SunToolkit.realSync(SunToolkit.java:1402)
at java.desktop/java.awt.Robot.waitForIdle(Robot.java:675)
at bug6980209.main(bug6980209.java:72)
at
java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
Method)
at
java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
at
java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.base/java.lang.reflect.Method.invoke(Method.java:564)
at
com.sun.javatest.regtest.agent.MainWrapper$MainThread.run(MainWrapper.java:127)
at java.base/java.lang.Thread.run(Thread.java:830)
- Exception reported in JDK-8198615, "It seems the secondary loop
will never end"
- Timeout in the realSync(), is it possible that we need to
increase the default timeout, or
we have a bug in realSync()?
BTW to check the intermittent test you need to run it more than 3
times(I guess 50 pre platform is fine).
On 11/19/19 1:42 am, Prasanta Sadhukhan wrote:
Hi All,
Please review a testbug fix for an issue seen intermittently in
linux nightly.
Issue is that the test times out after the test finishes without
actually propagating the test results to jtreg harness. It seems
like the JFrame which is created is called with EXIT_ON_CLOSE which
will call System.exit(0) causing the vm to terminate and if the
test is not run with othervm option, it might create some issue
when run in a bunch, as it is done in nightly testing.
Proposed fix is to remove the EXIT_ON_CLOSE option as we already
have frame.dispose() call. mach5 run is green.
Bug: https://bugs.openjdk.java.net/browse/JDK-8234385
diff -r 9db62a092725
test/jdk/java/awt/EventQueue/6980209/bug6980209.java
--- a/test/jdk/java/awt/EventQueue/6980209/bug6980209.java Thu Nov
14 10:59:45 2019 +0530
+++ b/test/jdk/java/awt/EventQueue/6980209/bug6980209.java Tue Nov
19 15:06:03 2019 +0530
@@ -63,7 +63,6 @@
public void run() {
frame = new JFrame();
frame.setUndecorated(true);
*- frame.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);*
setup(frame);
}
});
Regards
Prasanta
