<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 - __builtin_nontemporal_store spoils vectorization"
href="https://bugs.llvm.org/show_bug.cgi?id=46382">46382</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>__builtin_nontemporal_store spoils vectorization
</td>
</tr>
<tr>
<th>Product</th>
<td>clang
</td>
</tr>
<tr>
<th>Version</th>
<td>trunk
</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>-New Bugs
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedclangbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>jed@59a2.org
</td>
</tr>
<tr>
<th>CC</th>
<td>htmldeveloper@gmail.com, llvm-bugs@lists.llvm.org, neeilans@live.com, richard-llvm@metafoo.co.uk
</td>
</tr></table>
<p>
<div>
<pre>Use of Clang's __builtin_nontemporal_store() spoils vectorization on
clang-10/trunk:
void copy(int n, const double *a, double *b) {
for (int i = 0; i < n; i++)
if (1) {
__builtin_nontemporal_store(a[i], &b[i]);
} else {
b[i] = a[i];
}
}
<a href="https://gcc.godbolt.org/z/E-zCLZ">https://gcc.godbolt.org/z/E-zCLZ</a>
This generates movnti (scalar code), while the alternative gives movupd/movups.
There appears to be no way to obtain vmovntpd except via _mm*_stream_pd
intrinsics. This also relates to OpenMP-5's pragma omp simd nontemporal, which
(when implemented) would provide a portable alternative.
Nontemporal stores can provide a significant performance boost for
bandwidth-intensive applications. On my 2-socket EPYC 7452 node, for example,
it's the difference between 200 GB/s and 300 GB/s in STREAM, much of which
carries over to linear solves using Krylov methods.</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>