<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Good to know, thanks.  <br>
    <br>
    My comment wasn't specific to the patch in question.  I was trying
    to make a general point.  Sorry if my wording didn't convey that.  <br>
    <br>
    Philip<br>
    <br>
    <div class="moz-cite-prefix">On 05/05/2014 10:07 AM, David Majnemer
      wrote:<br>
    </div>
    <blockquote
cite="mid:CAL7bZ_eNSXSE6OopeCNWMkrycUpHgrbYenYnNRHTtH1k74enUA@mail.gmail.com"
      type="cite">Philip,
      <div><br>
      </div>
      <div>The patch submitted for review has already been updated to
        use Lint instead of Verifier.</div>
      <div><br>
      </div>
      <div>-- </div>
      <div>David Majnemer<br>
        <br>
        On Monday, May 5, 2014, Philip Reames <<a
          moz-do-not-send="true" href="mailto:listmail@philipreames.com">listmail@philipreames.com</a>>
        wrote:<br>
        <blockquote class="gmail_quote" style="margin:0 0 0
          .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <div bgcolor="#FFFFFF" text="#000000"> Speaking as someone
            with a language frontend which fairly frequently generates
            nonsensical code you'd never see out of Clang, I *strongly*
            agree that this functionality should not be in the
            Verifier.  :)<br>
            <br>
            Philip<br>
            <br>
            <div>On 05/03/2014 11:36 PM, Chandler Carruth wrote:<br>
            </div>
            <blockquote type="cite">
              <div dir="ltr">Heh. Was actua0lly looking at this in my
                post-commit review queue, and had the same comment. I
                think most of the inalloca stuff is *much* better
                fitting as a lint check than verifier.</div>
              <div class="gmail_extra"> <br>
                <br>
                <div class="gmail_quote">On Sat, May 3, 2014 at 10:55
                  PM, Chris Lattner <span dir="ltr"><<a
                      moz-do-not-send="true"
                      href="javascript:_e(%7B%7D,'cvml','clattner@apple.com');"
                      target="_blank">clattner@apple.com</a>></span>
                  wrote:<br>
                  <blockquote class="gmail_quote" style="margin:0 0 0
                    .8ex;border-left:1px #ccc solid;padding-left:1ex">
                    <div><br>
                      On May 2, 2014, at 3:28 PM, Reid Kleckner <<a
                        moz-do-not-send="true"
                        href="javascript:_e(%7B%7D,'cvml','rnk@google.com');"
                        target="_blank">rnk@google.com</a>> wrote:<br>
                      <br>
                      > So, we don't generally fail verification
                      because somebody gave us well-formed but
                      nonsensical IR.<br>
                      <br>
                    </div>
                    Right.<br>
                    <div><br>
                      > I kind of compromised on the last verifier
                      change because I think it will save us lots of
                      time, and we can revert it if somebody objects.
                       Consider the FAQ about calling convention
                      mismatches:<br>
                      ><br>
                      > <a moz-do-not-send="true"
href="http://llvm.org/docs/FAQ.html#why-does-instcombine-simplifycfg-turn-a-call-to-a-function-with-a-mismatched-calling-convention-into-unreachable-why-not-make-the-verifier-reject-it"
                        target="_blank">http://llvm.org/docs/FAQ.html#why-does-instcombine-simplifycfg-turn-a-call-to-a-function-with-a-mismatched-calling-convention-into-unreachable-why-not-make-the-verifier-reject-it</a><br>
                      ><br>
                      > It's not clear to me that this new check will
                      find lots of optimizer bugs for us, but I could be
                      proven wrong, and yes, it might take more time to
                      figure it out without a verifier check.  Anyway, I
                      think we need a second opinion to keep going in
                      this direction, so I added Nick.<br>
                      <br>
                    </div>
                    We really really should not put this kind of check
                    into the verifier.  The compiler needs to be able to
                    compile nonsensical but structurally valid code,
                    because it may be dynamically dead, and it may be
                    introduced by other transformations.<br>
                    <br>
                    Not all hope is lost though, this is a great thing
                    to add to llvm/lib/Analysis/Lint.cpp<br>
                    <span><font color="#888888"><br>
                        -Chris<br>
                      </font></span>
                    <div>
                      <div>_______________________________________________<br>
                        llvm-commits mailing list<br>
                        <a moz-do-not-send="true"
                          href="javascript:_e(%7B%7D,'cvml','llvm-commits@cs.uiuc.edu');"
                          target="_blank">llvm-commits@cs.uiuc.edu</a><br>
                        <a moz-do-not-send="true"
                          href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits"
                          target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits</a><br>
                      </div>
                    </div>
                  </blockquote>
                </div>
                <br>
              </div>
              <br>
              <fieldset></fieldset>
              <br>
              <pre>_______________________________________________
llvm-commits mailing list
<a moz-do-not-send="true" href="javascript:_e(%7B%7D,'cvml','llvm-commits@cs.uiuc.edu');" target="_blank">llvm-commits@cs.uiuc.edu</a>
<a moz-do-not-send="true" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits</a>
</pre>
            </blockquote>
            <br>
          </div>
        </blockquote>
      </div>
    </blockquote>
    <br>
  </body>
</html>