<html>
<head>
<base href="https://llvm.org/bugs/" />
</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 --- - Performance regression in AA"
href="https://llvm.org/bugs/show_bug.cgi?id=26564">26564</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>Performance regression in AA
</td>
</tr>
<tr>
<th>Product</th>
<td>new-bugs
</td>
</tr>
<tr>
<th>Version</th>
<td>3.8
</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>Keywords</th>
<td>performance
</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>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>bsteinbr@gmail.com
</td>
</tr>
<tr>
<th>CC</th>
<td>chandlerc@gmail.com, llvm-bugs@lists.llvm.org
</td>
</tr>
<tr>
<th>Classification</th>
<td>Unclassified
</td>
</tr></table>
<p>
<div>
<pre>Since upgrading to a more recent LLVM version for the Rust compiler, there has
been a large increase in compile time spent in LLVM optimizations. The
corresponding Rust bug report can be found here:
<a href="https://github.com/rust-lang/rust/issues/31435">https://github.com/rust-lang/rust/issues/31435</a>
At least part of that seems to come from the new AA handling, which exhibits
O(n^2) behavior in some cases with n being the number of active AA passes.
This originates from using the AAResultProxy in the AAResultBase class. For
example, calling AAResults::getModRefBehavior(CallSite) will iterate over all
registered AA results which all may eventually call into
AAResultBase::getModRefBehavior(CallSite) which calls
AAResultProxy::getModRefBehavior(Function*) which will then iterate over all
registered passes again.</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>