Uploaded image for project: 'JDK'
  1. JDK
  2. JDK-8252219

C2: Randomize IGVN worklist for stress testing

    XMLWordPrintable

Details

    • Enhancement
    • Resolution: Fixed
    • P4
    • 16
    • 16
    • hotspot
    • b18

    Description

      Bugs often depend on the order in which nodes are processed by IGVN. For example, IGVN worklist order depends on (incremental) inlining decisions that vary from compilation to compilation due to different profiling information which makes some bugs very hard to reproduce.

      Randomizing IGVN worklist access might help to trigger some bugs more often:
      - Node* n = _worklist.pop();
      + // Randomize access to worklist
      + int idx = (os::random() % _worklist.size());
      + Node* n = _worklist.remove(idx);

      I gave above code a quick try and executed tier1 - tier5: No new failures but it may still pay off to introduce this for stress testing.

      Attachments

        Issue Links

          relates to

          Bug - A problem which impairs or prevents the functions of the product. JDK-8253748 StressIGV tests fail with release VMs

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Resolved

          Bug - A problem which impairs or prevents the functions of the product. JDK-8253766 compiler/debug/TestGenerateStressSeed.java fails with NoSuchFileException

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. JDK-8253767 compiler/debug/TestStressIGVN.java fails with RuntimeException: got the same IGVN trace for different seeds: expected to not equal

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. JDK-8254810 gtest for os::random() fails on windows x64

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Enhancement - null JDK-8325482 Test that distinct seeds produce distinct traces for compiler stress flags

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Open

          Enhancement - null JDK-8253765 C2: Control randomization in StressLCM and StressGCM

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Resolved

          Enhancement - null JDK-8272703 StressSeed should be set via FLAG_SET_ERGO

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Resolved

          Enhancement - null CODETOOLS-7902768 jcstress should support -XX:+StressIGVN

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Enhancement - null CODETOOLS-7903041 jcstress: Provide explicit StressSeed when available

          • P4 - Minor loss of function, or other problem where easy workaround is present.
          • Closed

          Sub-task - The sub-task of the issue JDK-8314836 4 compiler/debug tests ignore tested VM flags

          • P5 - Cosmetic problem like misspelt words or misaligned text.
          • Open

          Bug - A problem which impairs or prevents the functions of the product. JDK-8311279 TestStressIGVNAndCCP.java failed with different IGVN traces for the same seed

          • P5 - Cosmetic problem like misspelt words or misaligned text.
          • Resolved
          links to

          Commit Commit openjdk/jdk/fed3636f

          Review Review openjdk/jdk/242

          Activity

            People

              rcastanedalo Roberto Castaneda Lozano
              thartmann Tobias Hartmann
              Votes:
              0 Vote for this issue
              Watchers:
              7 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: