<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Yes, and the only use case at the moment is for streaming to the
    IdealGraphVisualizer.<br>
    <br>
    //N<br>
    <br>
    Christian Thalinger skrev 2012-04-17 18:13:
    <blockquote
      cite="mid:83C33F83-E1D8-4E03-A375-DD099D7786A1@oracle.com"
      type="cite"><br>
      <div>
        <div>On Apr 13, 2012, at 1:21 AM, Nils Eliasson wrote:</div>
        <br class="Apple-interchange-newline">
        <blockquote type="cite">
          <meta content="text/html; charset=ISO-8859-1"
            http-equiv="Content-Type">
          <div bgcolor="#FFFFFF" text="#000000"> Can only find julong.
            u_long is defined in winsock.h that this is a call to.<br>
            <br>
            Changed the pointer cast to a correct construct.<br>
            <br>
            <a moz-do-not-send="true"
              href="http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.04/">http://cr.openjdk.java.net/~neliasso/7157695/webrev.04/</a><br>
            <br>
            Otherwise the same.<br>
          </div>
        </blockquote>
        <div><br>
        </div>
        This seems to be the least intrusive change and since nobody
        ever needed this before but you (and I guess you tested it) this
        looks good.</div>
      <div><br>
      </div>
      <div>-- Chris</div>
      <div><br>
        <blockquote type="cite">
          <div bgcolor="#FFFFFF" text="#000000"> <br>
            Thanks,<br>
            //Nils<br>
            <br>
            <br>
            Vladimir Kozlov skrev 2012-04-10 17:08:
            <blockquote cite="mid:4F844CD1.7040300@oracle.com"
              type="cite">+&nbsp; int ret = ::ioctlsocket(fd, FIONREAD,
              (u_long*)pbytes); <br>
              <br>
              What type is u_long? VM has defined type 'ulong' not
              'u_long'. <br>
              <br>
              Otherwise looks good. <br>
              <br>
              Vladimir <br>
              <br>
              On 4/10/12 6:45 AM, Nils Eliasson wrote: <br>
              <blockquote type="cite">Any comments on the latest rev? <br>
                <br>
                //Nils <br>
                <br>
                Nils Eliasson skrev 2012-03-30 10:40: <br>
                <blockquote type="cite">ok, <br>
                  <br>
                  New webrev here: <a moz-do-not-send="true"
                    class="moz-txt-link-freetext"
                    href="http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.03/">http://cr.openjdk.java.net/~neliasso/7157695/webrev.03/</a>
                  <br>
                  <a moz-do-not-send="true"
                    class="moz-txt-link-rfc2396E"
                    href="http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.03/">&lt;http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.03/&gt;</a>
                  <br>
                  <br>
                  Defining socklen_t as int on windows makes a much
                  better impl. <br>
                  <br>
                  Remains some casts from size_t to int on windows. So
                  we have jni call in jvm.h with jint/int -&gt; os.hpp
                  with size_t -&gt; <br>
                  os_windows.hpp with int. But we don't expect any
                  negative lenghts so it should work. <br>
                  <br>
                  Thanks for the feedback, <br>
                  Nils E <br>
                  <br>
                  David Holmes skrev 2012-03-29 23:20: <br>
                  <blockquote type="cite">On 30/03/2012 4:36 AM, Dean
                    Long wrote: <br>
                    <blockquote type="cite">Isn't this webrev reversing
                      the recent changes for 7091417? <br>
                    </blockquote>
                    <br>
                    Yes - and it shouldn't be. size_t's and socklen_t's
                    are the correct types to use. <br>
                    <br>
                    David <br>
                    <br>
                    <blockquote type="cite">dl <br>
                      <br>
                      On 3/29/2012 10:42 AM, Christian Thalinger wrote:
                      <br>
                      <blockquote type="cite">I'm pretty sure this gets
                        us into trouble with some compiler somewhere <br>
                        (I'd guess GCC on Linux). On Mac OS X send is
                        declared as: <br>
                        <br>
                        ssize_t <br>
                        send(int socket, const void *buffer, size_t
                        length, int flags); <br>
                        <br>
                        but length is now passed as int: <br>
                        <br>
                        -inline int os::send(int fd, char* buf, size_t
                        nBytes, uint flags) { <br>
                        +inline int os::send(int fd, char* buf, int
                        nBytes, int flags) { <br>
                        RESTARTABLE_RETURN_INT(::send(fd, buf, nBytes,
                        flags)); <br>
                        } <br>
                        <br>
                        I think your changes are good but we might need
                        some casts down in the <br>
                        OS files eventually. <br>
                        <br>
                        -- Chris <br>
                        <br>
                        On Mar 29, 2012, at 3:00 PM, Nils Eliasson
                        wrote: <br>
                        <br>
                        <blockquote type="cite"><a
                            moz-do-not-send="true"
                            class="moz-txt-link-freetext"
                            href="http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.01/">http://cr.openjdk.java.net/~neliasso/7157695/webrev.01/</a><a
                            moz-do-not-send="true"
                            class="moz-txt-link-rfc2396E"
                            href="http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.01/">&lt;http://cr.openjdk.java.net/%7Eneliasso/7157695/webrev.01/&gt;</a>
                          <br>
                          <br>
                          <br>
                          7157695 - Add windows implementation of socket
                          interface <br>
                          <br>
                          Using winsock api which is almost posix
                          compatible. Also changing <br>
                          some argument types for the socket interface
                          calls in os.hpp to make <br>
                          the platform independent code more platform
                          independent. This made <br>
                          the win implementation possible without casts,
                          and allowed me to <br>
                          remove a lot of redundant casts from the jni
                          exports in jvm.cpp. <br>
                        </blockquote>
                      </blockquote>
                    </blockquote>
                  </blockquote>
                  <br>
                </blockquote>
                <br>
                <!-- This signature was generated by the MyDesktop Oracle Business Signature utility version 3.8.8 -->
              </blockquote>
            </blockquote>
          </div>
        </blockquote>
      </div>
    </blockquote>
    <br>
  </body>
</html>