<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 - Compiler fails to optimize big endian load since clang 8.0"
   href="https://bugs.llvm.org/show_bug.cgi?id=43146">43146</a>
          </td>
        </tr>

        <tr>
          <th>Summary</th>
          <td>Compiler fails to optimize big endian load since clang 8.0
          </td>
        </tr>

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

        <tr>
          <th>Version</th>
          <td>8.0
          </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>normal
          </td>
        </tr>

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

        <tr>
          <th>Component</th>
          <td>LLVM Codegen
          </td>
        </tr>

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

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

        <tr>
          <th>CC</th>
          <td>llvm-bugs@lists.llvm.org, neeilans@live.com, richard-llvm@metafoo.co.uk
          </td>
        </tr></table>
      <p>
        <div>
        <pre>The compiler fails to recognize a BigEndian load opportunity since clang 7

Consider:

struct BESize {
    unsigned char c1;
    unsigned char c2;
    unsigned char c3;
    unsigned char c4;
    unsigned char c5;
    unsigned char c6;
    unsigned char c7;
    unsigned char c8;

    uint64_t Load() const {
        return 
            (static_cast<uint64_t>(c1) << 56) | 
            (static_cast<uint64_t>(c2) << 48) | 
            (static_cast<uint64_t>(c3) << 40) | 
            (static_cast<uint64_t>(c4) << 32) | 
            (static_cast<uint64_t>(c5) << 24) | 
            (static_cast<uint64_t>(c6) << 16) | 
            (static_cast<uint64_t>(c7) <<  8) | 
            static_cast<uint64_t>(c8);
    }
};

Compiling this with clang 7.0 -O3 -march=haswell, it creates proper movbe code:
As per <a href="https://gcc.godbolt.org/z/YklsAm">https://gcc.godbolt.org/z/YklsAm</a>

uint64_t load_be2(const BESize* p) {
    return p->Load();
}

-->

load_be2(BESize const*):                   # @load_be2(BESize const*)
        movbe   rax, qword ptr [rdi]
        ret


Starting with clang 8.0, it generates a bunch of avx code, see
<a href="https://gcc.godbolt.org/z/7rfh9J">https://gcc.godbolt.org/z/7rfh9J</a>


clang 8.0 still recognizes it is a Big Endian load, as running without
-march=haswell, it creates:

load_be2(BESize const*):                   # @load_be2(BESize const*)
        mov     rax, qword ptr [rdi]
        bswap   rax
        ret


Is this by design? I would guess the movbe is desirable over the AVX code?</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>