<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 - Pointless loop unroll / vectorization"
   href="https://bugs.llvm.org/show_bug.cgi?id=38280">38280</a>
          </td>
        </tr>

        <tr>
          <th>Summary</th>
          <td>Pointless loop unroll / vectorization
          </td>
        </tr>

        <tr>
          <th>Product</th>
          <td>libraries
          </td>
        </tr>

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

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

        <tr>
          <th>OS</th>
          <td>Windows NT
          </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>Loop Optimizer
          </td>
        </tr>

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

        <tr>
          <th>Reporter</th>
          <td>fabiang@radgametools.com
          </td>
        </tr>

        <tr>
          <th>CC</th>
          <td>llvm-bugs@lists.llvm.org
          </td>
        </tr></table>
      <p>
        <div>
        <pre>Example C++ code for x86, simplified from a more complex use case:

// ---- begin

#include <stdint.h>
#include <stddef.h>
#include <emmintrin.h>

// neg_offs <= -8 required
void apply_delta(uint8_t *dst, const uint8_t *src, ptrdiff_t neg_offs, size_t
count)
{
    // Just provided for context
    while (count >= 8)
    {
        __m128i src_bytes = _mm_loadl_epi64((const __m128i *) src);
        __m128i pred_bytes = _mm_loadl_epi64((const __m128i *) (dst +
neg_offs));
        __m128i sum = _mm_add_epi8(src_bytes, pred_bytes);
        _mm_storel_epi64((__m128i *) dst, sum);

        dst += 8;
        src += 8;
        count -= 8;
    }

    // This is the loop in question
    while (count--)
    {
        *dst = *src + dst[neg_offs];
        dst++;
        src++;
    }
}

// ---- end

The bottom (tail) loop gets expanded into a giant monstrosity that attempts to
process 64 bytes at once, with various special-case paths for tail processing,
to handle cases where neg_offs > -64 (which means the obvious
64-elements-at-a-time loop would not work), etc.

The full code can be viewed at <a href="https://godbolt.org/g/yRThcs">https://godbolt.org/g/yRThcs</a>, I won't post it
here. :)

All of which is completely pointless because the tail loop will (as is easy to
see) only ever see count <= 7.

This is an extreme example, but I'm seeing this general pattern (a scalar tail
loop for a manually vectorized loop getting pointlessly auto-vectorized) a lot.</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>