<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Great, thanks Jess<br>
    <br>
    -Dave<br>
    <br>
    <div class="moz-cite-prefix">On 7/22/15 4:23 PM, Calciano, Jess
      wrote:<br>
    </div>
    <blockquote
cite="mid:E079A211AAA8AD48953B67A89DB71C8533169134@fmsmsx104.amr.corp.intel.com"
      type="cite">
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <meta name="Generator" content="Microsoft Word 15 (filtered
        medium)">
      <style><!--
/* Font Definitions */
@font-face
        {font-family:"Cambria Math";
        panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
        {font-family:Calibri;
        panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
        {margin:0in;
        margin-bottom:.0001pt;
        font-size:12.0pt;
        font-family:"Times New Roman",serif;}
a:link, span.MsoHyperlink
        {mso-style-priority:99;
        color:blue;
        text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
        {mso-style-priority:99;
        color:purple;
        text-decoration:underline;}
p
        {mso-style-priority:99;
        mso-margin-top-alt:auto;
        margin-right:0in;
        mso-margin-bottom-alt:auto;
        margin-left:0in;
        font-size:12.0pt;
        font-family:"Times New Roman",serif;}
span.EmailStyle18
        {mso-style-type:personal-reply;
        font-family:"Calibri",sans-serif;
        color:#1F497D;}
.MsoChpDefault
        {mso-style-type:export-only;
        font-size:10.0pt;
        font-family:"Calibri",sans-serif;}
@page WordSection1
        {size:8.5in 11.0in;
        margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
        {page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
      <div class="WordSection1">
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Hello
            Dave,<o:p></o:p></span></p>
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Both
            reports look good and are approved to publish to the Logo
            List.<o:p></o:p></span></p>
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">--
            Jess<o:p></o:p></span></p>
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
        <div
          style="mso-element:para-border-div;border:none;border-bottom:solid
          windowtext 1.0pt;padding:0in 0in 1.0pt 0in">
          <p class="MsoNormal" style="border:none;padding:0in"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
        </div>
        <p class="MsoNormal"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
        <p class="MsoNormal"><a moz-do-not-send="true"
            name="_MailEndCompose"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></a></p>
        <p class="MsoNormal"><b><span
              style="font-size:11.0pt;font-family:"Calibri",sans-serif">From:</span></b><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif">
            David Wyman [<a class="moz-txt-link-freetext" href="mailto:dwyman@iol.unh.edu">mailto:dwyman@iol.unh.edu</a>]
            <br>
            <b>Sent:</b> Wednesday, July 22, 2015 12:08 PM<br>
            <b>To:</b> Calciano, Jess<br>
            <b>Cc:</b> <a class="moz-txt-link-abbreviated" href="mailto:iwg-arbitration-committee@openfabrics.org">iwg-arbitration-committee@openfabrics.org</a>; OFA
            Lab Mailing List; Cole, Cliff<br>
            <b>Subject:</b> Re: Updated arbitration request for Intel
            QLE7340 & QLE7342 HCAs (Jan 2015 OFA Interop Logo Event)<o:p></o:p></span></p>
        <p class="MsoNormal"><o:p> </o:p></p>
        <div>
          <p class="MsoNormal">Hello Jess,<o:p></o:p></p>
          <div>
            <p class="MsoNormal"><o:p> </o:p></p>
          </div>
          <div>
            <p class="MsoNormal">Your arbitration request regarding the
              NFSoRDMA and RSockets issues found in the January 2015
              Interop Logo Event has been approved and an updated report
              for Intel Infiniband HCAs is attached.  I also attached
              the Intel Switch Report.  Please review both and if you
              are satisfied with the content respond with explicit
              consent to post these reports to the OpenFabrics
              Interoperability Logo List.  Please contact me with any
              questions or concerns.<o:p></o:p></p>
          </div>
          <div>
            <p class="MsoNormal"><o:p> </o:p></p>
          </div>
          <div>
            <p class="MsoNormal">Thank you,<o:p></o:p></p>
          </div>
          <div>
            <p class="MsoNormal">Dave Wyman<o:p></o:p></p>
          </div>
          <div>
            <p class="MsoNormal">UNH-IOL OpenFabrics Interoperability
              Logo Group<o:p></o:p></p>
          </div>
          <div>
            <p class="MsoNormal"><o:p> </o:p></p>
          </div>
          <div>
            <p class="MsoNormal"><o:p> </o:p></p>
          </div>
          <div>
            <p class="MsoNormal"><o:p> </o:p></p>
            <div>
              <p class="MsoNormal"><o:p> </o:p></p>
              <div>
                <p class="MsoNormal"><o:p> </o:p></p>
                <div>
                  <p class="MsoNormal">On Wed, Jun 3, 2015 at 3:10 PM,
                    Calciano, Jess <<a moz-do-not-send="true"
                      href="mailto:jess.calciano@intel.com"
                      target="_blank">jess.calciano@intel.com</a>>
                    wrote:<o:p></o:p></p>
                  <blockquote style="border:none;border-left:solid
                    #CCCCCC 1.0pt;padding:0in 0in 0in
6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-bottom:5.0pt">
                    <div>
                      <div>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Hello,</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Since
                            the original arbitration request was
                            submitted, there’s been some further
                            discussion about the RSockets failure. With
                            the fix for librdmacm described in the
                            original request, rstream ran successfully
                            for most message sizes, but still hung with
                            -S 1024.</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Additional
                            investigation traced the new problem to an
                            incompatibility between the qib driver and
                            the ibv_create_qp() function. A workaround
                            (described below) is available for the
                            current OFED version and a permanent fix to
                            librdmacm will be included in the next OFED
                            3.18 release.</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Details:</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;margin-bottom:12.0pt;margin-left:.5in">
                          The ultimate issue is still related to the qib
                          driver being non-compliant with the
                          ibv_create_qp() definition:<br>
                          <br>
                                      The function ibv_create_qp() will
                          update the qp_init_attr->cap<br>
                                      struct with the actual QP values
                          of the QP that was created;<br>
                                      *** the values will be greater
                          than or equal to the values requested. ***<br>
                                      <br>
                          Specifically, the qib driver will return an
                          inline size that is smaller than that
                          requested.  Rsockets has code to trap for
                          this, but the rsockets code looks like this:<br>
                          <br>
                                      inline_size = SOME_DEFAULT_LIKE_64<br>
                                      rs_init_bufs(...);<br>
                                      ...<br>
                                      rs_create_qp(...);<br>
                                      inline_size =
                          qp_cap->max_inline_size;<br>
                          <br>
                          The issue is that rs_init_bufs(), which
                          allocates the buffers and registers the
                          memory, uses the default inline size.  The net
                          result is that rsockets ends up referencing
                          memory that is outside of the registered
                          memory region when sending credit updates. 
                          The lost credit update is causing the hang
                          that you see.<br>
                          <br>
                          A quick check shows that I can move the
                          rs_init_bufs() call after the qp has been
                          created and have the test work.  You should
                          also be able to override the inline_size by
                          writing the value 0 into a config file.  This
                          will set the inline_size to 0 as the default.
                           To do this, you need to write a 0 into
                          /etc/rdma/rsocket/inline_default.  (The actual
                          path will depend on your configuration, so it
                          could be under /usr/etc/rdma/... for example.)
                           Updating the config file should work with the
                          current version.<br>
                          <br>
                          I will provide an update to the librdmacm to
                          handle this.  That update will find its way
                          into the 3.18 release.<o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Thanks,</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Jess
                            Calciano</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><a
                            moz-do-not-send="true"
                            name="14dbad57595896c8__MailEndCompose"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span></a><o:p></o:p></p>
                        <div>
                          <div style="border:none;border-top:solid
                            #E1E1E1 1.0pt;padding:3.0pt 0in 0in 0in">
                            <p class="MsoNormal"
                              style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><b><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif">From:</span></b><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif">
                                Calciano, Jess
                                <br>
                                <b>Sent:</b> Wednesday, April 08, 2015
                                2:39 PM<br>
                                <b>To:</b> </span><a
                                moz-do-not-send="true"
                                href="mailto:iwg-arbitration-committee@openfabrics.org"
                                target="_blank"><span
                                  style="font-size:11.0pt;font-family:"Calibri",sans-serif">iwg-arbitration-committee@openfabrics.org</span></a><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif"><br>
                                <b>Cc:</b> OFA Lab Mailing List; Dave
                                Wyman; Rupert Dance <</span><a
                                moz-do-not-send="true"
                                href="mailto:rsdance@soft-forge.com"
                                target="_blank"><span
                                  style="font-size:11.0pt;font-family:"Calibri",sans-serif">rsdance@soft-forge.com</span></a><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif">>
                                (</span><a moz-do-not-send="true"
                                href="mailto:rsdance@soft-forge.com"
                                target="_blank"><span
                                  style="font-size:11.0pt;font-family:"Calibri",sans-serif">rsdance@soft-forge.com</span></a><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif">);
                                Cole, Cliff; Mascarenhas, Edward;
                                Sharma, Karun; Thete, Swapna; Hefty,
                                Sean; Yan, Philip W; Flores, Jose F<br>
                                <b>Subject:</b> Arbitration request for
                                Intel QLE7340 & QLE7342 HCAs (Jan
                                2015 OFA Interop Logo Event)</span><o:p></o:p></p>
                          </div>
                        </div>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"> <o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Hello,</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Intel
                            would like to file an arbitration request
                            for the January 2015 OFA Interop Logo Event
                            results for the Intel QLE7340 and QLE7342
                            HCAs.</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">The
                            provided report (attached for reference)
                            shows two failing tests:
                          </span><o:p></o:p></p>
                        <p><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">1)</span><span
                            style="font-size:7.0pt;color:#1F497D">     
                          </span><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">TI
                            NFS over RDMA</span><o:p></o:p></p>
                        <p><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">2)</span><span
                            style="font-size:7.0pt;color:#1F497D">     
                          </span><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">TI
                            RSockets</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">The
                            Intel team has investigated these results
                            and determined that the failures are due to
                            bugs in non-Intel components.</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">NFSoRDMA:</span><o:p></o:p></p>
                        <p class="MsoNormal"
style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;margin-left:.25in"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">The
                            failure is due to a known Connectathon
                            issue, documented here:</span><o:p></o:p></p>
                        <p class="MsoNormal"
style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto;margin-left:.25in"><a
                            moz-do-not-send="true"
                            href="http://www.spinics.net/lists/linux-nfs/msg16460.html"
                            target="_blank"><span
                              style="font-size:11.0pt;font-family:"Calibri",sans-serif">http://www.spinics.net/lists/linux-nfs/msg16460.html</span></a><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">RSockets:</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;margin-bottom:12.0pt;margin-left:.25in">
                          <span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#002060">The
                            issue is that ibv_modify_qp() is failing. 
                            The problem is that an incorrect bit is set
                            in the qp_attr_mask, which is returned from
                            the kernel.  With Intel, bit 21 of the
                            qp_attr_mask is set.  This is not the case
                            for a Mellanox HCA.<br>
                            <br>
                            Bit 21 is not defined for userspace. 
                            However, it was defined in the kernel as
                            IB_QP_SMAC.<br>
                            <br>
                            If the librdmacm is modified to mask out
                            this bit, the call succeeds and rstream runs
                            successfully.</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Please
                            let me know if the arbitration committee
                            needs any additional information on the
                            analysis.</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"> </span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Thanks,</span><o:p></o:p></p>
                        <p class="MsoNormal"
                          style="mso-margin-top-alt:auto;mso-margin-bottom-alt:auto"><span
style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Jess
                            Calciano</span><o:p></o:p></p>
                      </div>
                    </div>
                  </blockquote>
                </div>
                <p class="MsoNormal"><o:p> </o:p></p>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
  </body>
</html>