<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 12/12/2017 7:26 AM, Sean Silva via

      llvm-dev wrote:<br>

    </div>

    <blockquote type="cite"

cite="mid:CAHnXoakxA7AU+8ZAR9PiW9H18tR8hORW=rKs8SMBwfNtJ0Rhyg@mail.gmail.com">

      <div dir="auto">

        <div><br>

          <div class="gmail_extra"><br>

            <div class="gmail_quote">On Dec 11, 2017 6:54 PM, "Sean

              Silva" <<a href="mailto:chisophugis@gmail.com"

                moz-do-not-send="true">chisophugis@gmail.com</a>>

              wrote:<br type="attribution">

              <blockquote class="quote" style="margin:0 0 0

                .8ex;border-left:1px #ccc solid;padding-left:1ex">

                <div dir="ltr">

                  <div>It looks like in various places in SelectionDAG

                    and related machinery we appear to implicitly or

                    explicitly assume that TypeExpandInteger only

                    applies to scalars.<br>

                  </div>

                  <div><br>

                    Does anybody know why that is?</div>

                  <div><br>

                  </div>

                  <div>For our target (Pixel Visual Core: <a

href="https://www.blog.google/products/pixel/pixel-visual-core-image-processing-and-machine-learning-pixel-2/"

                      target="_blank" moz-do-not-send="true">https://www.blog.google/<wbr>products/pixel/pixel-visual-<wbr>core-image-processing-and-<wbr>machine-learning-pixel-2/</a>)

                    expanding a N-bit vector op into two (N/2)-bit

                    vector ops</div>

                </div>

              </blockquote>

            </div>

          </div>

        </div>

        <div dir="auto"><br>

        </div>

        <div dir="auto">To be clear: the vector ops are all of the same

          number of elements. Just the bit width of the element type

          changes.</div>

      </div>

    </blockquote>

    <br>

    I think it just hasn't been a priority for any in-tree target; most

    code doesn't use illegal vector types much, and the default strategy

    generates decent code in most cases.  (You might have an easier time

    just adding a few target-specific DAGCombines for specific

    operations which lead to low-quality code.)<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>