<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <br>
    <div class="moz-cite-prefix">On 2015-01-15 23:00, Jungwoo Ha wrote:<br>
    </div>
    <blockquote
cite="mid:CA+n_jhh8vtr5-=c4Rq7hnzt+Vq91btVSt_4xcRY8fxoQcup=Tw@mail.gmail.com"
      type="cite">
      <div dir="ltr">I agree with not having a flag. It seems like an
        obvious improvement.</div>
    </blockquote>
    <br>
    Sound fair.<br>
    <br>
    Latest patch looks goo to me too.<br>
    <br>
    I'll sponsor this push.<br>
    <br>
    Bengt<br>
    <br>
    <blockquote
cite="mid:CA+n_jhh8vtr5-=c4Rq7hnzt+Vq91btVSt_4xcRY8fxoQcup=Tw@mail.gmail.com"
      type="cite">
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Thu, Jan 15, 2015 at 1:58 PM, Kim
          Barrett <span dir="ltr"><<a moz-do-not-send="true"
              href="mailto:kim.barrett@oracle.com" target="_blank">kim.barrett@oracle.com</a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div class="HOEnZb">
              <div class="h5">On Jan 15, 2015, at 12:00 PM, Jungwoo Ha
                <<a moz-do-not-send="true"
                  href="mailto:jwha@google.com">jwha@google.com</a>>
                wrote:<br>
                ><br>
                > Apparently, my memory's been swapped out. I meant
                to use _promotion_failed field.<br>
                ><br>
                > diff -r a184ee1d7172
                src/share/vm/gc_implementation/parNew/parNewGeneration.cpp<br>
                > ---
                a/src/share/vm/gc_implementation/parNew/parNewGeneration.cpp 
                    Thu Jan 08 12:08:22 2015 -0800<br>
                > +++
                b/src/share/vm/gc_implementation/parNew/parNewGeneration.cpp 
                    Thu Jan 15 09:00:05 2015 -0800<br>
                > @@ -1194,8 +1194,10 @@<br>
                >          return real_forwardee(old);<br>
                >      }<br>
                ><br>
                > -    new_obj =
                _next_gen->par_promote(par_scan_state->thread_num(),<br>
                > -                                       old, m,
                sz);<br>
                > +    if (!_promotion_failed) {<br>
                > +      new_obj =
                _next_gen->par_promote(par_scan_state->thread_num(),<br>
                > +                                        old, m,
                sz);<br>
                > +    }<br>
                ><br>
                >      if (new_obj == NULL) {<br>
                >        // promotion failed, forward to self<br>
                <br>
              </div>
            </div>
            Yes, this looks right.<br>
            <br>
            Bengt mentioned protection with a flag from the original
            proposed change.  I’m not sure there’s a need for a new flag
            for this version of the change.<br>
            <br>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </body>
</html>