Reduce lock contention in Thread.getAndClearInterrupt

[babsingh]

Lock contention in Thread.getAndClearInterrupt is reduced by
acquiring the lock only when the value of the Thread.interrupted
field is true.

Related: eclipse-openj9/openj9#20414

[tajila]

jenkins test sanity alinux64 jdknext

[babsingh]

In the PR builds, the following known failures were seen:

• jdk24 java/lang/Thread/virtual/MonitorWaitNotify AssertionFailedError: expected: <true> but was: <false> eclipse-openj9/openj9#20369
• jdk24 java.lang.AssertionError: sun.security.util.ResourcesMgr (PD) eclipse-openj9/openj9#20351
• jdk24 java/foreign/handles/Driver Command-line option unrecognised: --illegal-native-access=deny eclipse-openj9/openj9#20354
• jdk24 java/foreign/largestub/TestLargeStub IllegalArgumentException array length is not legal for long[] or double[] eclipse-openj9/openj9#20355

[babsingh]

Two undocumented failures are seen (see below for more details). I don't feel that these are caused due to this PR. These failures are seen probably because the JDK-next repo is unstable.

• TestMappedHandshake: The assert related to accessExecutor.awaitTermination has failed due to the timeout. This indicates that the threads need more time to run. Ran a 25x grinder: https://openj9-jenkins.osuosl.org/job/Grinder/3964. The test didn't fail in the grinder.

10:05:43  STDOUT:
10:05:43  test TestMappedHandshake.testHandshake(): failure
10:05:43  java.lang.AssertionError: expected [true] but found [false]
10:05:43  	at org.testng.Assert.fail(Assert.java:99)
10:05:43  	at org.testng.Assert.failNotEquals(Assert.java:1037)
10:05:43  	at org.testng.Assert.assertTrue(Assert.java:45)
10:05:43  	at org.testng.Assert.assertTrue(Assert.java:55)
10:05:43  	at TestMappedHandshake.testHandshake(TestMappedHandshake.java:90)

• KDFDelayedProviderSyncTest: Another timeout. Ran a 25x grinder: https://openj9-jenkins.osuosl.org/job/Grinder/3965. This one looks intermittent.

09:43:14  test KDFDelayedProviderSyncTest.testDerive(): failure
09:43:14  org.testng.internal.thread.ThreadTimeoutException: Method KDFDelayedProviderSyncTest.testDerive() didn't finish within the time-out 150
09:43:14  	at org.testng.internal.MethodInvocationHelper.invokeWithTimeoutWithNoExecutor(MethodInvocationHelper.java:322)
09:43:14  	at org.testng.internal.MethodInvocationHelper.invokeWithTimeout(MethodInvocationHelper.java:280)
09:43:14  	at org.testng.internal.TestInvoker.invokeMethod(TestInvoker.java:605)
09:43:14  	at org.testng.internal.TestInvoker.invokeTestMethod(TestInvoker.java:174)
09:43:14  	at org.testng.internal.MethodRunner.runInSequence(MethodRunner.java:46)
09:43:14  	at org.testng.internal.TestInvoker$MethodInvocationAgent.invoke(TestInvoker.java:822)
09:43:14  	at org.testng.internal.TestInvoker.invokeTestMethods(TestInvoker.java:147)
09:43:14  	at org.testng.internal.TestMethodWorker.invokeTestMethods(TestMethodWorker.java:146)
09:43:14  	at org.testng.internal.TestMethodWorker.run(TestMethodWorker.java:128)
09:43:14  	at org.testng.internal.thread.ThreadUtil.lambda$execute$0(ThreadUtil.java:66)
09:43:14  	at java.base/java.util.concurrent.FutureTask.run(FutureTask.java:328)
09:43:14  	at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1144)
09:43:14  	at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:642)
09:43:14  	at java.base/java.lang.Thread.run(Thread.java:1588)

[babsingh]

OpenJ9 issues to track the two new failures:

• jdk24 javax/crypto/KDF/KDFDelayedProviderSyncTest ThreadTimeoutException: didn't finish within the time-out 150 eclipse-openj9/openj9#20526
• jdk24 java/foreign/TestMappedHandshake AssertionError: expected [true] but found [false] eclipse-openj9/openj9#20527