Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8172015

ConcurrentQueues/OfferRemoveLoops.java fails

    Details

    • Type: Bug
    • Status: Open
    • Priority: P4
    • Resolution: Unresolved
    • Affects Version/s: 9
    • Fix Version/s: tbd_minor
    • Component/s: core-libs
    • Labels:

      Description

      After merging the latest Hotspot changes and Dev java/util/concurrent/ConcurrentQueues/OfferRemoveLoops.java started failing (on all platforms):

      ----------messages:(4/142)----------
      command: main OfferRemoveLoops 300
      reason: User specified action: run main OfferRemoveLoops 300
      Mode: othervm
      elapsed time (seconds): 51.895
      ----------configuration:(0/0)----------
      ----------System.out:(3/30)----------
      Passed = 983672, failed = 1

      ----------System.err:(24/1061)----------
      LinkedBlockingQueue
      LinkedBlockingQueue
      LinkedBlockingDeque
      LinkedBlockingDeque
      ArrayBlockingQueue
      PriorityBlockingQueue
      ConcurrentLinkedDeque
      ConcurrentLinkedQueue
      Hung thread: remover
      LinkedTransferQueue
      java.lang.AssertionError: Some tests failed
      at OfferRemoveLoops.instanceMain(OfferRemoveLoops.java:187)
      at OfferRemoveLoops.main(OfferRemoveLoops.java:183)
      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:538)
      at com.sun.javatest.regtest.agent.MainWrapper$MainThread.run(MainWrapper.java:110)
      at java.base/java.lang.Thread.run(Thread.java:844)

      JavaTest Message: Test threw exception: java.lang.AssertionError: Some tests failed
      JavaTest Message: shutting down test

      STATUS:Failed.`main' threw exception: java.lang.AssertionError: Some tests failed

        Activity

        Hide
        martin Martin Buchholz added a comment -
        As typically happens, I cannot reproduce this failure (on linux-amd64). Perhaps it's actually a hotspot bug due to the recent integration? Please try harder to characterize the failures and bisect to the failing commit. One possibility is the recent AOT work, which my jdk isn't using (due to missing libelf).
        Show
        martin Martin Buchholz added a comment - As typically happens, I cannot reproduce this failure (on linux-amd64). Perhaps it's actually a hotspot bug due to the recent integration? Please try harder to characterize the failures and bisect to the failing commit. One possibility is the recent AOT work, which my jdk isn't using (due to missing libelf).
        Hide
        martin Martin Buchholz added a comment -
        Testing team needs to do more work to isolate the problem, so assigning back to Dmitry.
        Show
        martin Martin Buchholz added a comment - Testing team needs to do more work to isolate the problem, so assigning back to Dmitry.
        Hide
        dholmes David Holmes added a comment -
        All we have is:

        Hung thread: remover

        which basically indicates a timeout - the remover thread had not terminated by the time the main thread was done waiting for completion. But there is no stack dump for the hung thread

        if (thread.isAlive()) {
          System.err.printf("Hung thread: %s%n", thread.getName());
          failed++;
          for (StackTraceElement e : thread.getStackTrace())
            System.err.println(e);
         thread.interrupt();
        }

        How can that be?

        The test execution parameters were:

          -Xcomp -Xcomp -XX:MaxRAMFraction=8 -XX:+CreateCoredumpOnCrash -ea -esa -XX:CompileThreshold=100 -XX:+UnlockExperimentalVMOptions -server -XX:-TieredCompilation -XX:+IgnoreUnrecognizedVMOptions -XX:+DeoptimizeALot

        which may suggest a compiler deopt issue.
        Show
        dholmes David Holmes added a comment - All we have is: Hung thread: remover which basically indicates a timeout - the remover thread had not terminated by the time the main thread was done waiting for completion. But there is no stack dump for the hung thread if (thread.isAlive()) {   System.err.printf("Hung thread: %s%n", thread.getName());   failed++;   for (StackTraceElement e : thread.getStackTrace())     System.err.println(e);  thread.interrupt(); } How can that be? The test execution parameters were:   -Xcomp -Xcomp -XX:MaxRAMFraction=8 -XX:+CreateCoredumpOnCrash -ea -esa -XX:CompileThreshold=100 -XX:+UnlockExperimentalVMOptions -server -XX:-TieredCompilation -XX:+IgnoreUnrecognizedVMOptions -XX:+DeoptimizeALot which may suggest a compiler deopt issue.
        Hide
        psandoz Paul Sandoz added a comment -
        Such a failure has not been observed on mach5 indicating this is not urgent (so downgrading to P4), nor can i reproduce locally with the "exotic" HotSpot flags, but such failures are still occurring in our system running with those "exotic" flags.

        This needs some focused effort to run the test repeatedly on an identified machine to sufficiently reproduce, then point a HotSpot engineer at it to debug.
        Show
        psandoz Paul Sandoz added a comment - Such a failure has not been observed on mach5 indicating this is not urgent (so downgrading to P4), nor can i reproduce locally with the "exotic" HotSpot flags, but such failures are still occurring in our system running with those "exotic" flags. This needs some focused effort to run the test repeatedly on an identified machine to sufficiently reproduce, then point a HotSpot engineer at it to debug.
        Hide
        martin Martin Buchholz added a comment -
        We can't reproduce this failure.
        Show
        martin Martin Buchholz added a comment - We can't reproduce this failure.

          People

          • Assignee:
            dfazunen Dmitry Fazunenko
            Reporter:
            dfazunen Dmitry Fazunenko
          • Votes:
            0 Vote for this issue
            Watchers:
            4 Start watching this issue

            Dates

            • Created:
              Updated: