<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Hi everyone!<br>
    <br>
    Since the decision was made to change only tests which fail because
    of conflict for now (skip "selfish" tests), I post new webrev for
    jdk part of the <a moz-do-not-send="true" id="key-val"
      rel="4684019"
      href="https://bugs.openjdk.java.net/browse/JDK-8019361">JDK-8019361</a>:<br>
    <a class="moz-txt-link-freetext" href="http://cr.openjdk.java.net/~avstepan/eistepan/8062536/webrev.03/">http://cr.openjdk.java.net/~avstepan/eistepan/8062536/webrev.03/</a><br>
    <br>
    Thanks,<br>
    Evgeniya Stepanova
    <div class="moz-cite-prefix">On 07.11.2014 15:34, Evgeniya Stepanova
      wrote:<br>
    </div>
    <blockquote cite="mid:545CAE30.9010708@oracle.com" type="cite">
      <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
      David, Filipp, Katja<br>
      <br>
      Diff have been updated one more time:<br>
      java/lang/management/RuntimeMXBean/TestInputArgument.sh and
      test/java/lang/ref/EnqueuePollRace.java have been changed<br>
      <a moz-do-not-send="true" class="moz-txt-link-freetext"
        href="http://cr.openjdk.java.net/%7Eeistepan/8062536/webrev.02/">http://cr.openjdk.java.net/~eistepan/8062536/webrev.02/</a><br>
      <br>
      Thanks<br>
      <br>
      <div class="moz-cite-prefix">On 07.11.2014 9:37, David Holmes
        wrote:<br>
      </div>
      <blockquote cite="mid:545C5AB6.6030103@oracle.com" type="cite">On
        7/11/2014 12:36 AM, Evgeniya Stepanova wrote: <br>
        <blockquote type="cite">New webrev: <br>
          <a moz-do-not-send="true" class="moz-txt-link-freetext"
            href="http://cr.openjdk.java.net/%7Eeistepan/8062536/webrev.01/">http://cr.openjdk.java.net/~eistepan/8062536/webrev.01/</a>
          <br>
        </blockquote>
        <br>
        In: <br>
        <br>
        test/java/lang/management/RuntimeMXBean/TestInputArgument.sh <br>
        <br>
        the use of the gc options seems incidental - it's just picking
        two innocuous options to use - similar to the JpsHelper case.
        You could replace +UseParallelGC with something like
        +UseFastJNIAccessors (platform independent and normally true). <br>
        <br>
        David <br>
        ----- <br>
        <br>
        <blockquote type="cite">Thanks, <br>
          Evgeniya Stepanova <br>
          On 06.11.2014 17:35, Yekaterina Kantserova wrote: <br>
          <blockquote type="cite">Thanks a lot! <br>
            <br>
            On 11/06/2014 02:05 PM, Evgeniya Stepanova wrote: <br>
            <blockquote type="cite">Hi Katja, <br>
              <br>
              Ok, this seems to be a perfect solution. Thank you. I'll
              change the <br>
              diff accordingly. <br>
              <br>
              <br>
              Thanks, <br>
              Evgeniya Stepanova <br>
              On 06.11.2014 16:56, Yekaterina Kantserova wrote: <br>
              <blockquote type="cite">Hi Dima, <br>
                <br>
                On 11/06/2014 11:22 AM, Dmitry Fazunenko wrote: <br>
                <blockquote type="cite">Hi Katja, <br>
                  <br>
                  You are right, there will be no conflict, because test
                  ignores any <br>
                  external VM flags. <br>
                  So, adding @requires seems unnecessary here, but... <br>
                  <br>
                  Ignoring external options is bad thing, such "selfish"
                  tests are <br>
                  not applicable for other areas, like GC, compiler, RT.
                  <br>
                </blockquote>
                <br>
                This particular case is to test the defined flags are
                shown up as <br>
                expected. <br>
                <br>
                Evgeniya, <br>
                <br>
                would you mind to change JpsHelper.java instead? <br>
                <br>
                +++ b/test/sun/tools/jps/JpsHelper.java <br>
                @@ -93,7 +93,7 @@ <br>
                     /** <br>
                      * VM arguments to start test application with <br>
                      */ <br>
                -    public static final String[] VM_ARGS = {"-Xmx512m",
                <br>
                "-XX:+UseParallelGC"}; <br>
                +    public static final String[] VM_ARGS = {"-Xmx512m",
                <br>
                "-XX:+PrintGCDetails"}; <br>
                     /** <br>
                      * VM flag to start test application with <br>
                      */ <br>
                <br>
                Best regards, <br>
                Katja <br>
                <br>
                <br>
                <br>
                <blockquote type="cite"> <br>
                  @requires  will allow to modify tests to include
                  external vm <br>
                  options without any risk of bumping into conflict and
                  extend area <br>
                  of test applicability. <br>
                  <br>
                  But if you still believe, that @requires is not
                  necessary - it's <br>
                  not a problem, tests could be kept as is. <br>
                  <br>
                  Thanks, <br>
                  Dima <br>
                  <br>
                  <br>
                  On 06.11.2014 16:27, Yekaterina Kantserova wrote: <br>
                  <blockquote type="cite"> <br>
                    Hi Evgeniya, <br>
                    <br>
                    As David has pointed out these jps tests are not
                    testing gc. The <br>
                    -XX:+UseParallelGC is just an arbitrary chosen test
                    flag. There <br>
                    should not be any conflicts either since these tests
                    are running <br>
                    in driver mode: <br>
                    <br>
                    ... <br>
                     * @run driver TestJpsJar <br>
                    ... <br>
                    <br>
                    which means no flags from above are accepted. <br>
                    <br>
                    Thanks, <br>
                    Katja <br>
                    <br>
                    <br>
                    <br>
                    On 11/06/2014 11:05 AM, Evgeniya Stepanova wrote: <br>
                    <blockquote type="cite">Hi David, <br>
                      <br>
                      tag added because tests contain string <br>
                       cmd.addAll(JpsHelper.getVmArgs()); <br>
                      <br>
                      and JpsHelper defines <br>
                      ... <br>
                      public static final String[] VM_ARGS =
                      {"-Xmx512m", <br>
                      "-XX:+UseParallelGC"}; <br>
                      ... <br>
                      public static List<String> getVmArgs()
                      throws IOException { <br>
                              if (testVmArgs == null) { <br>
                                  testVmArgs = new ArrayList<>();
                      <br>
                                 
                      testVmArgs.addAll(Arrays.asList(VM_ARGS)); <br>
                                  testVmArgs.add("-XX:Flags=" + <br>
                      getVmFlagsFile().getAbsolutePath()); <br>
                              } <br>
                              return testVmArgs; <br>
                          } <br>
                      <br>
                      Tests itself wouldn't fail if we use another GC,
                      tag added for <br>
                      cleanup-if we use for example SerialGC we must be
                      sure that tests <br>
                      passed with this GC, not with another one. Now you
                      will assume <br>
                      that concrete test passed with Serial GC, but it
                      run only with <br>
                      Parallel GC. Plus there is no any sense to run
                      test twice in TC <br>
                      (with different GC, since it use only Parallel) <br>
                      <br>
                      Thanks, <br>
                      Evgeniya Stepanova <br>
                      On 06.11.2014 6:20, David Holmes wrote: <br>
                      <blockquote type="cite">Hi Evgeniya, <br>
                        <br>
                        On 6/11/2014 1:33 AM, Evgeniya Stepanova wrote:
                        <br>
                        <blockquote type="cite">Hi, <br>
                          <br>
                          Please review changes for 8062536, the
                          OpenJDK/jdk part of the <br>
                          JDK-8019361 <br>
                          <br>
                          bug: <a moz-do-not-send="true"
                            class="moz-txt-link-freetext"
                            href="https://bugs.openjdk.java.net/browse/JDK-8062536">https://bugs.openjdk.java.net/browse/JDK-8062536</a>
                          <br>
                          fix: <a moz-do-not-send="true"
                            class="moz-txt-link-freetext"
                            href="http://cr.openjdk.java.net/%7Eeistepan/8062536/webrev.00/">http://cr.openjdk.java.net/~eistepan/8062536/webrev.00/</a>
                          <br>
                          <br>
                          Problem: Some tests explicitly set GC and fail
                          when another GC <br>
                          is set <br>
                          outside <br>
                        </blockquote>
                        <br>
                        I don't see why you have done this for the <br>
                        <br>
                        test/sun/tools/jps/TestJps*.java <br>
                        <br>
                        tests. They don't set any GC flags. <br>
                        <br>
                        <blockquote type="cite">Solution: Such tests
                          marked with the jtreg tag "requires" to <br>
                          skip test <br>
                          if there is a conflict <br>
                        </blockquote>
                        <br>
                        Just wondering: Does a skipped test get a .jtr
                        file showing it <br>
                        was skipped; or does it only appear in the
                        higher-level jtreg log? <br>
                        <br>
                        Thanks, <br>
                        David <br>
                        <br>
                        <blockquote type="cite">Tested locally with
                          different GC flags (-XX:+UseG1GC, <br>
                          -XX:+UseParallelGC, -XX:+UseSerialGC,
                          -XX:+UseConcMarkSweep and <br>
                          without <br>
                          any GC flag). Tests are being excluded as
                          expected. No tests <br>
                          failed <br>
                          because of the conflict. <br>
                          <br>
                          Thanks, <br>
                          Evgeniya Stepanova <br>
                          <br>
                          // <br>
                        </blockquote>
                      </blockquote>
                      <br>
                      -- <br>
                      /Evgeniya Stepanova/ <br>
                    </blockquote>
                    <br>
                    <br>
                    <br>
                  </blockquote>
                  <br>
                </blockquote>
                <br>
              </blockquote>
              <br>
              -- <br>
              /Evgeniya Stepanova/ <br>
            </blockquote>
            <br>
          </blockquote>
          <br>
          -- <br>
          /Evgeniya Stepanova/ <br>
        </blockquote>
      </blockquote>
      <br>
      <div class="moz-signature">-- <br>
        <i>Evgeniya Stepanova</i></div>
    </blockquote>
    <br>
    <div class="moz-signature">-- <br>
      <i>Evgeniya Stepanova</i></div>
  </body>
</html>