<html>
    <head>
      <base href="https://bugs.llvm.org/">
    </head>
    <body><table border="1" cellspacing="0" cellpadding="8">
        <tr>
          <th>Bug ID</th>
          <td><a class="bz_bug_link 
          bz_status_NEW "
   title="NEW - Bad for loop copy optimization when using -fno-builtin-memcpy -fno-builtin-memmove"
   href="https://bugs.llvm.org/show_bug.cgi?id=38103">38103</a>
          </td>
        </tr>

        <tr>
          <th>Summary</th>
          <td>Bad for loop copy optimization when using -fno-builtin-memcpy -fno-builtin-memmove
          </td>
        </tr>

        <tr>
          <th>Product</th>
          <td>clang
          </td>
        </tr>

        <tr>
          <th>Version</th>
          <td>6.0
          </td>
        </tr>

        <tr>
          <th>Hardware</th>
          <td>PC
          </td>
        </tr>

        <tr>
          <th>OS</th>
          <td>Linux
          </td>
        </tr>

        <tr>
          <th>Status</th>
          <td>NEW
          </td>
        </tr>

        <tr>
          <th>Severity</th>
          <td>enhancement
          </td>
        </tr>

        <tr>
          <th>Priority</th>
          <td>P
          </td>
        </tr>

        <tr>
          <th>Component</th>
          <td>C++
          </td>
        </tr>

        <tr>
          <th>Assignee</th>
          <td>unassignedclangbugs@nondot.org
          </td>
        </tr>

        <tr>
          <th>Reporter</th>
          <td>gchatelet@google.com
          </td>
        </tr>

        <tr>
          <th>CC</th>
          <td>dgregor@apple.com, llvm-bugs@lists.llvm.org
          </td>
        </tr></table>
      <p>
        <div>
        <pre>Straightforward copy using a for loop with known size at compile time leads to
very poor assembly.

<span class="quote">> #include <cstddef>

> template <size_t kBlockSize>
> void Copy(char* __restrict dst, const char* __restrict src) {
>   for (size_t i = 0; i < kBlockSize; ++i) dst[i] = src[i];
> }

> template void Copy<15>(char* __restrict dst, const char* __restrict src);</span >

<a href="https://godbolt.org/g/YFq3o6">https://godbolt.org/g/YFq3o6</a>

This can be mitigated by the introduction of a temporary buffer like so:
<span class="quote">> template <size_t kBlockSize>
> void Copy(char* __restrict dst, const char* __restrict src) {
>   char tmp[kBlockSize];
>   for (size_t i = 0; i < kBlockSize; ++i) tmp[i] = src[i];
>   for (size_t i = 0; i < kBlockSize; ++i) dst[i] = tmp[i];
> }</span >

<a href="https://godbolt.org/g/48Dghk">https://godbolt.org/g/48Dghk</a>

It works up to 25B and produce bad code from 26B onwards.

Check the resulting code for 32B for instance: <a href="https://godbolt.org/g/jZwcrv">https://godbolt.org/g/jZwcrv</a></pre>
        </div>
      </p>


      <hr>
      <span>You are receiving this mail because:</span>

      <ul>
          <li>You are on the CC list for the bug.</li>
      </ul>
    </body>
</html>