Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8279196

Test: jdk/jfr/event/gc/stacktrace/TestG1OldAllocationPendingStackTrace.java timed out

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: P4 P4
    • tbd
    • 11.0.21-oracle, 18, 19, 20, 23
    • hotspot
    • jfr

      The test passes but then times out during the test cleanup. Normally this tests runs in about 12s on Windows but in this case we run 480+ seconds before getting the event we are looking for:

      [488.120s][info][gc ] GC(136586) Pause Young (Normal) (G1 Evacuation Pause) 64M->61M(128M) 9.639ms
      [488.120s][info][gc,cpu ] GC(136586) User=0.08s Sys=0.00s Real=0.02s
      JFR GC events found: 124501
      Event: jdk.AllocationRequiringGC {
        startTime = 05:30:44.703 (2021-12-23)
        gcId = 2
        size = 10.0 kB
        eventThread = "MainThread" (javaThreadId = 16)
        stackTrace = [
          jdk.jfr.event.gc.stacktrace.OldGenMemoryAllocator.allocate() line: 87
          jdk.jfr.event.gc.stacktrace.AllocationStackTrace.allocAndCheck(GarbageCollectorMXBean, MemoryAllocator, String[]) line: 377
          jdk.jfr.event.gc.stacktrace.AllocationStackTrace.testAllocEvent(GarbageCollectorMXBean, MemoryAllocator, String[]) line: 343
          jdk.jfr.event.gc.stacktrace.AllocationStackTrace.testG1OldAllocEvent() line: 283
          jdk.jfr.event.gc.stacktrace.TestG1OldAllocationPendingStackTrace.main(String[]) line: 37
          ...
        ]
      }


      [488.171s][info][gc,heap,exit ] Heap
      [488.171s][info][gc,heap,exit ] garbage-first heap total 131072K, used 71377K [0x00000000f8000000, 0x0000000100000000)
      [488.171s][info][gc,heap,exit ] region size 1024K, 9 young (9216K), 2 survivors (2048K)
      [488.171s][info][gc,heap,exit ] Metaspace used 6374K, committed 6592K, reserved 1114112K
      [488.171s][info][gc,heap,exit ] class space used 648K, committed 768K, reserved 1048576K
      ----------System.err:(1/16)----------
      STATUS:Passed.

      This test failed in a CI run that included the fix for JDK-8278125 which changes the number of stackful OOME instances that can be observed. As the failing test mentions OOME and stacktrace they may be related - though I cannot immediately see how.

            mgronlun Markus Grönlund
            dholmes David Holmes
            Votes:
            0 Vote for this issue
            Watchers:
            6 Start watching this issue

              Created:
              Updated: