<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">On 5/25/2018 11:46 AM, JF Bastien
      wrote:<br>
    </div>
    <blockquote type="cite"
      cite="mid:7A085E79-CC0C-4709-8405-2F2032DFDC3F@apple.com">
      <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
      <br class="">
      <div><br class="">
        <blockquote type="cite" class="">
          <div class="">On May 25, 2018, at 11:38 AM, Friedman, Eli <<a
              href="mailto:efriedma@codeaurora.org" class=""
              moz-do-not-send="true">efriedma@codeaurora.org</a>>
            wrote:</div>
          <br class="Apple-interchange-newline">
          <div class="">
            <meta http-equiv="Content-Type" content="text/html;
              charset=utf-8" class="">
            <div text="#000000" bgcolor="#FFFFFF" class="">
              <div class="moz-cite-prefix">On 5/25/2018 11:29 AM, JF
                Bastien via cfe-dev wrote:<br class="">
              </div>
              <blockquote type="cite"
                cite="mid:AC7AF159-592E-4EA7-9AA4-A8242D750526@apple.com"
                class="">
                <meta http-equiv="Content-Type" content="text/html;
                  charset=utf-8" class="">
                <div class="">
                  <div class="">
                    <ol class="MailOutline">
                      <li class="">Teach the target infrastructure that
                        hardware interference size is something they can
                        specify (in tablegen files somewhere).</li>
                      <li class="">Allow overriding the value in
                        sub-targets using -march or -mcpu (the
                        sub-target defines the numeric value, and the
                        user gets the overriden one by using -march or
                        -mcpu).</li>
                    </ol>
                  </div>
                </div>
              </blockquote>
              <br class="">
              We can't change the value based on -mcpu.  We generally
              allow mixing code built with different values of -mcpu. 
              And any code which is linked together must use the same
              value for hardware_destructive_interference_size, or else
              we violate ODR.<br class="">
            </div>
          </div>
        </blockquote>
        <div><br class="">
        </div>
        <div>Interesting point. The case I’d like to cover is one where
          the developer wants to get the exact right value for their
          particular CPU, instead of a conservative answer with extra
          padding. How do you think we should meet this use case?</div>
        <br class="">
      </div>
    </blockquote>
    <br>
    Go back to the standards committee and ask for a function that isn't
    constexpr?  I can't think of any other reasonable solution.<br>
    <br>
    -Eli<br>
    <pre class="moz-signature" cols="72">-- 
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a Linux Foundation Collaborative Project</pre>
  </body>
</html>