<html>
<head>
<base href="https://bugs.llvm.org/">
</head>
<body><span class="vcard"><a class="email" href="mailto:kubakuderski@gmail.com" title="Jakub Kuderski <kubakuderski@gmail.com>"> <span class="fn">Jakub Kuderski</span></a>
</span> changed
<a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - opt crashes with "opt -instcombine -adce -newgvn -gvn-hoist ": Assertion `DT->dominates(NewBB, OldBB) && "invalid path"' failed"
href="https://bugs.llvm.org/show_bug.cgi?id=34466">bug 34466</a>
<br>
<table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>What</th>
<th>Removed</th>
<th>Added</th>
</tr>
<tr>
<td style="text-align:right;">Resolution</td>
<td>---
</td>
<td>FIXED
</td>
</tr>
<tr>
<td style="text-align:right;">Assignee</td>
<td>unassignedbugs@nondot.org
</td>
<td>kubakuderski@gmail.com
</td>
</tr>
<tr>
<td style="text-align:right;">Status</td>
<td>NEW
</td>
<td>RESOLVED
</td>
</tr></table>
<p>
<div>
<b><a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - opt crashes with "opt -instcombine -adce -newgvn -gvn-hoist ": Assertion `DT->dominates(NewBB, OldBB) && "invalid path"' failed"
href="https://bugs.llvm.org/show_bug.cgi?id=34466#c13">Comment # 13</a>
on <a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED FIXED - opt crashes with "opt -instcombine -adce -newgvn -gvn-hoist ": Assertion `DT->dominates(NewBB, OldBB) && "invalid path"' failed"
href="https://bugs.llvm.org/show_bug.cgi?id=34466">bug 34466</a>
from <span class="vcard"><a class="email" href="mailto:kubakuderski@gmail.com" title="Jakub Kuderski <kubakuderski@gmail.com>"> <span class="fn">Jakub Kuderski</span></a>
</span></b>
<pre>I could not reproduce the bug with reduced test-cases with r314161 -- I tried
both the IR and c source version and don't get any ICEs. Only the original c
source triggers ICE for me.
-------------------------------------------------------------
This bug should be fixed by r314254. The problem was that the incremental
updater in ADCE would not invalidate DT's DFS numbers when it performed only
deletions (and no insertions). This must confuse some other transform and make
it misbehave, but I was not able to track it down precisely.
I think that a true 'fix' would be to make the DomTree verifier also check DFS
numbers if they are marked as valid.</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>