Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8263554

4 applications/microbenchmarks/api/Test_4.java sub-tests timeout out



    • Type: Bug
    • Status: New
    • Priority: P4
    • Resolution: Unresolved
    • Affects Version/s: 17
    • Fix Version/s: 18
    • Component/s: hotspot
    • Subcomponent:
    • CPU:
    • OS:


      The following tests timed out in the JDK17 CI:


      Here's a log file snippet from Test_4.java#id15:

      [jmh] # JMH version: 1.21
      [jmh] # VM version: JDK 17-ea, Java HotSpot(TM) 64-Bit Server VM, 17-ea+13-LTS-995
      [jmh] # VM invoker: c:\\ade\\mesos\\work_dir\\jib-master\\install\\jdk-17+13-995\\windows-x64-debug.jdk\\jdk-17\\fastdebug\\bin\\java.exe
      [jmh] # VM options: -XX:MaxRAMPercentage=6 -Djava.io.tmpdir=t:\\testoutput\\test-support\\jtreg_closed_test_hotspot_jtreg_applications_microbenchmarks_api_Test_4_java\\tmp -XX:+CreateCoredumpOnCrash -ea -esa -XX:CompileThreshold=100 -XX:+UnlockExperimentalVMOptions -XX:+TieredCompilation -XX:+VerifyOops --add-opens=java.base/java.io=ALL-UNNAMED -Djmh.ignoreLock=true -Djava.io.tmpdir=T:\\testoutput\\test-support\\jtreg_closed_test_hotspot_jtreg_applications_microbenchmarks_api_Test_4_java\\scratch\\1
      [jmh] # Warmup: 5 iterations, 10 s each
      [jmh] # Measurement: 5 iterations, 10 s each
      [jmh] # Timeout: 10 min per iteration
      [jmh] # Threads: 1 thread, will synchronize iterations
      [jmh] # Benchmark mode: Throughput, ops/time
      [jmh] # Benchmark: oracle.micro.benchmarks.api.java.lang.ArrayCopyUnalignedSrc.arrayCopyLongUnalignedSrc0005
      [jmh] # Run progress: 0.00% complete, ETA 00:01:40
      [jmh] # Fork: 1 of 1
      [jmh] # Warmup Iteration 1: 26942412.662 ops/s
      [jmh] # Warmup Iteration 2: 35290284.159 ops/s
      [jmh] # Warmup Iteration 3: 29177370.436 ops/s
      [jmh] # Warmup Iteration 4: 35821801.206 ops/s
      [jmh] # Warmup Iteration 5: 28823970.574 ops/s
      [jmh] Iteration 1: 39209327.184 ops/s
      [jmh] Iteration 2: 39655560.559 ops/s
      [jmh] Iteration 3: 40312488.556 ops/s
      [jmh] Iteration 4: 40347210.720 ops/s
      [jmh] Iteration 5: 20097879.218 ops/s
      [jmh] Result "oracle.micro.benchmarks.api.java.lang.ArrayCopyUnalignedSrc.arrayCopyLongUnalignedSrc0005":
      [jmh] 35924493.247 \u00b1(99.9%) 34117207.263 ops/s [Average]
      [jmh] (min, avg, max) = (20097879.218, 35924493.247, 40347210.720), stdev = 8860130.382
      [jmh] CI (99.9%): [1807285.985, 70041700.510] (assumes normal distribution)
      [jmh] # Run complete. Total time: 00:20:44
      [jmh] REMEMBER: The numbers below are just data. To gain reusable insights, you need to follow up on
      [jmh] why the numbers are the way they are. Use profilers (see -prof, -lprof), design factorial
      [jmh] experiments, perform baseline and negative tests that provide experimental control, make sure
      [jmh] the benchmarking environment is safe on JVM/OS/HW level, ask for reviews from the domain experts.
      [jmh] Do not assume the numbers tell you what you want them to tell.
      [jmh] Benchmark Mode Cnt Score Error Units
      [jmh] ArrayCopyUnalignedSrc.arrayCopyLongUnalignedSrc0005 thrpt 5 35924493.247 \u00b1 34117207.263 ops/s

      JavaTest Message: Test complete.

      result: Error. Agent error: java.lang.Exception: Agent 7 timed out with a timeout of 1200 seconds; check console log for any additional details

      Based on "JavaTest Message: Test complete.", it looks like the
      test passed, but then timed out. Normally when a test times out
      in the middle of a run, you have a line like this:

      Timeout refired 1200 times

      and that line appears in "#section:driver" so it looks like this
      test passed while the timeout handler was running. The other
      three sub-test failures have the same markers.

      I think this test might need a longer timeout value.

      The test task's JVM args are:
      -XX:+CreateCoredumpOnCrash -ea -esa -XX:CompileThreshold=100 -XX:+UnlockExperimentalVMOptions -server -XX:+TieredCompilation -XX:+VerifyOops




            Unassigned Unassigned
            dcubed Daniel Daugherty
            0 Vote for this issue
            1 Start watching this issue