<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">On 13/02/14 15:45, Jon Masamitsu wrote:<br>
    </div>
    <blockquote cite="mid:52FCDA91.6010808@oracle.com" type="cite">
      <meta content="text/html; charset=ISO-8859-1"
        http-equiv="Content-Type">
      <br>
      <div class="moz-cite-prefix">On 2/12/14 11:51 PM, Stefan Karlsson
        wrote:<br>
      </div>
      <blockquote cite="mid:52FC7989.2040603@oracle.com" type="cite">
        <meta content="text/html; charset=ISO-8859-1"
          http-equiv="Content-Type">
        <div class="moz-cite-prefix">On 13/02/14 03:14, Jon Masamitsu
          wrote:<br>
        </div>
        <blockquote cite="mid:52FC2A77.20403@oracle.com" type="cite">
          <meta content="text/html; charset=ISO-8859-1"
            http-equiv="Content-Type">
          Stefan,<br>
          <br>
          I understand now that the same work is being<br>
          done (maybe in a different place) and that<br>
          the&nbsp; amount of extra work being<br>
          done (the call to test_set_oops_do_mark())<br>
          will scale with the number of nmethods.&nbsp; <br>
          <br>
          Is that correct?<br>
        </blockquote>
        <br>
        Yes, that's correct.<br>
        <br>
      </blockquote>
      <br>
      Ok.&nbsp; <br>
      <br>
      Change looks good.&nbsp; Thanks for an explanation of<br>
      the performance measurements done.<br>
    </blockquote>
    <br>
    Thanks, Jon.<br>
    <br>
    StefanK<br>
    <blockquote cite="mid:52FCDA91.6010808@oracle.com" type="cite"> <br>
      Jon<br>
      <br>
      <blockquote cite="mid:52FC7989.2040603@oracle.com" type="cite">
        StefanK<br>
        <blockquote cite="mid:52FC2A77.20403@oracle.com" type="cite"> <br>
          Jon<br>
          <br>
          <div class="moz-cite-prefix">On 2/12/2014 11:40 AM, Jon
            Masamitsu wrote:<br>
          </div>
          <blockquote cite="mid:52FBCE32.2000708@oracle.com" type="cite">
            <meta content="text/html; charset=ISO-8859-1"
              http-equiv="Content-Type">
            <br>
            <div class="moz-cite-prefix">On 2/12/2014 1:43 AM, Stefan
              Karlsson wrote:<br>
            </div>
            <blockquote cite="mid:52FB4245.1030901@oracle.com"
              type="cite">
              <meta http-equiv="content-type" content="text/html;
                charset=ISO-8859-1">
              Hi all,<br>
              <br>
              Please, review this patch to remove the do_code_roots
              parameter from SharedHeap::process_strong_roots.<br>
              <br>
              The changes done are:<br>
              - Change the code to rely on the ScannningOption so
              parameter instead of do_code_roots.<br>
              - Change GenMarkSweep and G1MarkSweep to adjust the code
              roots with the help of process_strong_roots instead of
              doing it as a separate phase after process_strong_roots.<br>
              - Removed the unused
              <meta http-equiv="content-type" content="text/html;
                charset=ISO-8859-1">
              FalseClosure.<br>
              <br>
              After this patch the adjust phase of the GenMarkSweep and
              G1MarkSweep will use the generic code in
              process_strong_roots, which mark/claim the nmethods before
              they are processed. Before the patch these two Serial Old
              GC adjust phases skipped the mark/claim part. No
              noticeable Serial Old GC time increases were found when
              this patch was performance tested.<br>
            </blockquote>
            <br>
            Does this mean ("adjust phase ...") that the "mark/claim"&nbsp;
            does not&nbsp; have any affect<br>
            on later processing?&nbsp; Or actually does nothing (even though
            the closures<br>
            are applied)?&nbsp; Which benchmarks did you use for performance
            testing?<br>
            <br>
            Jon<br>
            <blockquote cite="mid:52FB4245.1030901@oracle.com"
              type="cite"> <br>
              This cleanup is needed/wanted for G1 Class Unloading.<br>
              <br>
              Webrev:<br>
              <a moz-do-not-send="true" class="moz-txt-link-freetext"
                href="http://cr.openjdk.java.net/%7Estefank/8034761/webrev.00/">http://cr.openjdk.java.net/~stefank/8034761/webrev.00/</a><br>
              <br>
              RFE:<br>
              <a moz-do-not-send="true" class="moz-txt-link-freetext"
                href="https://bugs.openjdk.java.net/browse/JDK-8034761">https://bugs.openjdk.java.net/browse/JDK-8034761</a><br>
              <br>
              thanks,<br>
              StefanK<br>
            </blockquote>
            <br>
          </blockquote>
          <br>
        </blockquote>
        <br>
      </blockquote>
      <br>
    </blockquote>
    <br>
  </body>
</html>