<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 - [X86][SSE] X86TTIImpl::getVectorInstrCost - improve extract/insert element costs"
href="https://bugs.llvm.org/show_bug.cgi?id=43605">43605</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>[X86][SSE] X86TTIImpl::getVectorInstrCost - improve extract/insert element costs
</td>
</tr>
<tr>
<th>Product</th>
<td>libraries
</td>
</tr>
<tr>
<th>Version</th>
<td>trunk
</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>Backend: X86
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>llvm-dev@redking.me.uk
</td>
</tr>
<tr>
<th>CC</th>
<td>craig.topper@gmail.com, llvm-bugs@lists.llvm.org, llvm-dev@redking.me.uk, spatel+llvm@rotateright.com
</td>
</tr></table>
<p>
<div>
<pre>The costs for vector element extraction/insertion are underestimated:
- Non/Unknown immediate index cases should be considered high cost
- Vectors before SSE41 only <8 x i16> has PINSR/PEXTR support (+MOVD/MOVQ from
index=0) - the others must be shuffled+extracted or inserted+shuffled.
- Account for the subvector handling for immediates from vector elements above
the bottom xmm.
- GPR<->XMM can be pretty nasty, assuming throughput cost = 1 is an
underestimate in most cases</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>