<table border="1" cellspacing="0" cellpadding="8">
    <tr>
        <th>Issue</th>
        <td>
            <a href=https://github.com/llvm/llvm-project/issues/54849>54849</a>
        </td>
    </tr>

    <tr>
        <th>Summary</th>
        <td>
            Explicit weight definition
        </td>
    </tr>

    <tr>
      <th>Labels</th>
      <td>
      </td>
    </tr>

    <tr>
      <th>Assignees</th>
      <td>
      </td>
    </tr>

    <tr>
      <th>Reporter</th>
      <td>
          deeprobin
      </td>
    </tr>
</table>

<pre>
    Currently it is possible to define by means of `cold` that an entry point is rather rarely called and by means of `hot` that it is a hot path.

Assuming I have meaningful statistics that tell me how often a method is called or how hot the path is, this could be useful for the edge weight calculation.

I would suggest introducing a weight attribute that takes a floating point parameter.

-1.0 = cold
0 = unknown
1.0 = hot

For ex. 0.7 is hot but not so hot as 1.0

What do you think of this proposal?
</pre>
<img width="1px" height="1px" alt="" src="http://email.email.llvm.org/o/eJxdU9tu2zAM_Rr7haghW07qPPghaxagX7Bn2aJtLYpk6NI0fz9KTrZ2QBCCInnO4cWDlff-LTqHJug7qADKw2q9V4NGCBYkTsogDHe4ojAe7ATFno1WSzIQFhFAGKBqd6c6ZTKAE2FBR8YhgY5Ca5SUJv-HWWz4i7JRC6A3WKm-KtipYMft_-h9vCozwzss4gMzCLlT1OCDCMoHNfoNJ6DWFCecG7EENIR5xbBYmfAfWqzL8cRFSjMfRYvmjdyUZaMmsQjRY-KYKD_loZwRbqjmJSSkMWrituab1He45Wof5xk9NUWjsTKOSb14FosQnBpiwIdmccHU-6QtAVLiNslVOEHS0X0jeKkrBgU_QV5Cftr8aC7G3sz29ExKI_5SfKZO8LMCVr2mcaQBkAwwZL3NrvBAtV9rfiWJ0sLdxjQec0nby3NanaVbEbrg51L2XB74QZRBBY39z89Vq5GW-ug435FK0yqj0_0SwuoLfiyaM_1mFZY4VKO9kqP1x9O8EMFvHAO5ii4AaUPnXdu1h3Lpa8YPfGibuuvEjgvWvbaSYbvfDS3iocFSiwG174vdj2J3KlXfsKZhbV3XvOasq9gk21feshoPdbefeNEyvAqlq0RcWTeXrs8ahjh7Cmo6Mv8vKOgTmQ3iE19EOjHXS0TSPChTZsF9VvsH2IEemA">