<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <br>
    On 2012-02-24 13:50, Jesper Wilhelmsson wrote:
    <blockquote
      cite="mid:5B839334-95F3-47DB-801F-609C924379E9@oracle.com"
      type="cite">
      <div>I can have a look at it but I don't have the time to do it
        today. If Monday is OK then I can take it.</div>
    </blockquote>
    <br>
    Thanks, Jesper!<br>
    <br>
    Will also need a second review from someone with OpenJDK reviewer
    status.<br>
    <br>
    Bengt<br>
    <br>
    <blockquote
      cite="mid:5B839334-95F3-47DB-801F-609C924379E9@oracle.com"
      type="cite">
      <div>/Jesper<br>
        <br>
        <br>
      </div>
      <div><br>
        24 feb 2012 kl. 13:14 skrev Bengt Rutisson <<a
          moz-do-not-send="true" href="mailto:bengt.rutisson@oracle.com">bengt.rutisson@oracle.com</a>>:<br>
        <br>
      </div>
      <blockquote type="cite">
        <div>
          <meta http-equiv="content-type" content="text/html;
            charset=UTF-8">
          <br>
          Hi all,<br>
          <br>
          Just pinging this review request. Does anybody have some time
          to look at it? It is a fairly small and straight forward
          change...<br>
          <br>
          Thanks,<br>
          Bengt<br>
          <br>
          -------- Original Message --------
          <table class="moz-email-headers-table" border="0"
            cellpadding="0" cellspacing="0">
            <tbody>
              <tr>
                <th align="RIGHT" nowrap="nowrap" valign="BASELINE">Subject:
                </th>
                <td>Request for review (S): 7068625 Testing 8 bytes of
                  card table entries at a time speeds up card-scanning</td>
              </tr>
              <tr>
                <th align="RIGHT" nowrap="nowrap" valign="BASELINE">Date:
                </th>
                <td>Tue, 21 Feb 2012 12:03:50 +0400</td>
              </tr>
              <tr>
                <th align="RIGHT" nowrap="nowrap" valign="BASELINE">From:
                </th>
                <td>Alexey Ragozin <a moz-do-not-send="true"
                    class="moz-txt-link-rfc2396E"
                    href="mailto:alexey.ragozin@gmail.com"><alexey.ragozin@gmail.com></a></td>
              </tr>
              <tr>
                <th align="RIGHT" nowrap="nowrap" valign="BASELINE">To:
                </th>
                <td><a moz-do-not-send="true"
                    class="moz-txt-link-abbreviated"
                    href="mailto:hotspot-gc-dev@openjdk.java.net">hotspot-gc-dev@openjdk.java.net</a></td>
              </tr>
              <tr>
                <th align="RIGHT" nowrap="nowrap" valign="BASELINE">CC:
                </th>
                <td>Bengt Rutisson <a moz-do-not-send="true"
                    class="moz-txt-link-rfc2396E"
                    href="mailto:bengt.rutisson@oracle.com"><bengt.rutisson@oracle.com></a></td>
              </tr>
            </tbody>
          </table>
          <br>
          <br>
          Hi,<br>
          <br>
          I would like few volunteers to review changes for <a
            moz-do-not-send="true"
            href="http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7068625"
            target="_blank">http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7068625</a><br>
          WebRev: <a moz-do-not-send="true"
            href="http://cr.openjdk.java.net/%7Ebrutisso/7068625/webrev.00/"
            target="_blank">http://cr.openjdk.java.net/~brutisso/7068625/webrev.00/</a><br>
          <br>
          Change summary<br>
          For large heaps (I was focusing on 8GiB and above) it is
          common to have long continuous ranges of clean cards.<br>
          Patch is introducing a short path for skipping ranges of clean
          cards using word aligned memory access instead of byte
          aligned.<br>
          <br>
          Patch affects serial and CMS collectors. For CMS collector
          stride size should be increase to see any performance gains (I
          was using <br>
          <div class="MsoNormal"
            style="line-height:normal;margin-bottom:.0001pt;margin-bottom:0cm">
            <span style="font-family:"Courier New"">-XX:+UnlockDiagnosticVMOptions

            </span></div>
          <div class="MsoNormal"
            style="line-height:normal;margin-bottom:.0001pt;margin-bottom:0cm"><span
              style="font-family:"Courier New"">-XX:ParGCCardsPerStrideChunk=4096)<br>
              <br>
            </span>For testing I was mainly using synthetic benchmark
            randomly modifying hash tables in heap, thus uniformly
            touching cards across heaps.<br>
            Average duration of young GC pause were used as KPI.<br>
            More details about testing can be found at <br>
            <a moz-do-not-send="true"
href="http://blog.ragozin.info/2011/07/openjdk-patch-cutting-down-gc-pause.html">http://blog.ragozin.info/2011/07/openjdk-patch-cutting-down-gc-pause.html</a><br>
            Though article is referring jdk6, my resent tests with trunk
            jdk7 show no difference.<br>
            I was also tested patch with real application (Oracle
            Coherence storage node).<br>
            With 16GiB of heap and CMS/ParNew GC, enabling patch have
            shortened GC pauses roughly in 2 times.<br>
            <br>
            Source code of benchmark used in test are available at<br>
            <a moz-do-not-send="true"
href="https://gridkit.googlecode.com/svn/branches/aragozin-sandbox/young-gc-bench">https://gridkit.googlecode.com/svn/branches/aragozin-sandbox/young-gc-bench</a><br>
            Main class YoungGCPauseBenchmark<br>
          </div>
          <br>
          Regards,<br>
          Alexey<br>
          <br>
        </div>
      </blockquote>
    </blockquote>
    <br>
  </body>
</html>