<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">On 09/13/2013 10:09 AM, Stefan Karlsson
      wrote:<br>
    </div>
    <blockquote cite="mid:5232C842.5010408@oracle.com" type="cite">
      <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
      <div class="moz-cite-prefix">On 09/13/2013 09:48 AM, Bernd
        Eckenfels wrote:<br>
      </div>
      <blockquote
        cite="mid:1F8E7AEF-B1A9-405E-99B7-C20052C5D169@eckenfels.net"
        type="cite">
        <div><span class="Apple-style-span"
            style="-webkit-tap-highlight-color: rgba(26, 26, 26,
            0.296875); -webkit-composition-fill-color: rgba(175, 192,
            227, 0.230469); -webkit-composition-frame-color: rgba(77,
            128, 180, 0.230469); ">
            <div>Question: why is the trace block testing for
              is_humongus itself?</div>
            <div><br>
            </div>
            <div>I would expect the allocation code to branch somewhere
              for this condition anyway - would it make more sense to
              put the trace there? Especially if you want to maintain
              (and output) additional details/statistics.</div>
          </span></div>
      </blockquote>
      <br>
      You are probably right, and I'm not going to defend the current
      placement of this trace code. But for this specific change I just
      want to fix the crash and the other misplaced print out.<br>
      <br>
      <blockquote
        cite="mid:1F8E7AEF-B1A9-405E-99B7-C20052C5D169@eckenfels.net"
        type="cite">
        <div><span class="Apple-style-span"
            style="-webkit-tap-highlight-color: rgba(26, 26, 26,
            0.296875); -webkit-composition-fill-color: rgba(175, 192,
            227, 0.230469); -webkit-composition-frame-color: rgba(77,
            128, 180, 0.230469); ">
            <div><br>
            </div>
            <div>Is PTR_FORMAT right for word_size()</div>
          </span></div>
      </blockquote>
      <br>
      No, it should be SIZE_FORMAT. I'll change it.<br>
    </blockquote>
    <br>
    On the other hand, maybe this is intentional. The other humongous
    allocation output also print with PTR_FORMAT.<br>
    <br>
    Example:<br>
    Metadata humongous allocation:<br>
    Â  word_size 0x0000000000002620<br>
    Â  chunk_word_size 0x0000000000002628<br>
    Â Â Â  chunk overhead 0x0000000000000008<br>
    Â  new humongous chunk word size 0x0000000000002628<br>
    <br>
    I'll leave this for now as well.<br>
    <br>
    StefanK<br>
    <br>
    <blockquote cite="mid:5232C842.5010408@oracle.com" type="cite"> <br>
      <blockquote
        cite="mid:1F8E7AEF-B1A9-405E-99B7-C20052C5D169@eckenfels.net"
        type="cite">
        <div><span class="Apple-style-span"
            style="-webkit-tap-highlight-color: rgba(26, 26, 26,
            0.296875); -webkit-composition-fill-color: rgba(175, 192,
            227, 0.230469); -webkit-composition-frame-color: rgba(77,
            128, 180, 0.230469); ">
            <div> and would it be better to output bytes?</div>
          </span></div>
      </blockquote>
      <br>
      In my opinion, yes. However,  currently, most of print-outs
      guarded by TraceMetadata output numbers in words instead of bytes.
      Changing that will require another changeset.<br>
      <br>
      <blockquote
        cite="mid:1F8E7AEF-B1A9-405E-99B7-C20052C5D169@eckenfels.net"
        type="cite">
        <div><span class="Apple-style-span"
            style="-webkit-tap-highlight-color: rgba(26, 26, 26,
            0.296875); -webkit-composition-fill-color: rgba(175, 192,
            227, 0.230469); -webkit-composition-frame-color: rgba(77,
            128, 180, 0.230469); ">
            <div> Is this also available as a event?</div>
          </span></div>
      </blockquote>
      <br>
      Not at the moment. But we plan to send this information out as an
      event. Maybe we can clean up some of the issues you have pointed
      when that is done.<br>
      <br>
      thanks,<br>
      StefanK<br>
      <br>
      <blockquote
        cite="mid:1F8E7AEF-B1A9-405E-99B7-C20052C5D169@eckenfels.net"
        type="cite">
        <div><span class="Apple-style-span"
            style="-webkit-tap-highlight-color: rgba(26, 26, 26,
            0.296875); -webkit-composition-fill-color: rgba(175, 192,
            227, 0.230469); -webkit-composition-frame-color: rgba(77,
            128, 180, 0.230469); ">
            <div><br>
            </div>
            <div>Bernd</div>
          </span><br>
          Am 13.09.2013 um 08:57 schrieb Stefan Karlsson <<a
            moz-do-not-send="true"
            href="mailto:stefan.karlsson@oracle.com">stefan.karlsson@oracle.com</a>>:<br>
          <br>
        </div>
        <blockquote type="cite">
          <div>
            <meta content="text/html; charset=UTF-8"
              http-equiv="Content-Type">
            <div class="moz-cite-prefix">On 09/13/2013 08:22 AM, Bengt
              Rutisson wrote:<br>
            </div>
            <blockquote cite="mid:5232AF1E.5030108@oracle.com"
              type="cite">
              <meta content="text/html; charset=UTF-8"
                http-equiv="Content-Type">
              <div class="moz-cite-prefix"><br>
                Hi Stefan,<br>
                <br>
                On 9/12/13 10:00 PM, Stefan Karlsson wrote:<br>
              </div>
              <blockquote cite="mid:52321D71.2070008@oracle.com"
                type="cite"><a moz-do-not-send="true"
                  class="moz-txt-link-freetext"
                  href="http://cr.openjdk.java.net/%7Estefank/8024751/webrev.00/">http://cr.openjdk.java.net/~stefank/8024751/webrev.00/</a>
                <br>
                <br>
                Small fixes two fix some issues when TraceMetadata*
                flags are turned on. <br>
                <br>
                - TraceMetadataHumongousAllocation crashes. <br>
                - TraceMetadataChunkAllocation prints the same
                block_freelist() multiple times. <br>
              </blockquote>
              <br>
              Looks good. I'm fine with pushing this as is, but I think
              I would have preferred that this code:<br>
              <br>
              <meta http-equiv="content-type" content="text/html;
                charset=UTF-8">
              <br>
              2369   if (next != NULL) {<br>
              2370     if (TraceMetadataHumongousAllocation &&<br>
              2371        
              SpaceManager::is_humongous(next->word_size())) {<br>
              2372       gclog_or_tty->print_cr("  new humongous
              chunk word size " PTR_FORMAT,<br>
              2373                              next->word_size());<br>
              2374     }<br>
              2375   }<br>
              <br>
              was more like:<br>
              <br>
              2370     if (TraceMetadataHumongousAllocation &&<br>
              2371         next != NULL &&
              SpaceManager::is_humongous(next->word_size())) {<br>
              2372       gclog_or_tty->print_cr("  new humongous
              chunk word size " PTR_FORMAT,<br>
              2373                              next->word_size());<br>
              2374     }<br>
              <br>
              To me it makes it clearer that this is only a tracing
              section.<br>
            </blockquote>
            <br>
            Fair enough. I'll change it.<br>
            <br>
            thanks,<br>
            StefanK<br>
            <blockquote cite="mid:5232AF1E.5030108@oracle.com"
              type="cite"> <br>
              Thanks,<br>
              Bengt<br>
              <blockquote cite="mid:52321D71.2070008@oracle.com"
                type="cite"> <br>
                thanks, <br>
                StefanK <br>
                <br>
                <br>
              </blockquote>
              <br>
            </blockquote>
            <br>
          </div>
        </blockquote>
      </blockquote>
      <br>
    </blockquote>
    <br>
  </body>
</html>