<html>
<head>
<base href="https://bugs.llvm.org/">
</head>
<body><span class="vcard"><a class="email" href="mailto:vsk@apple.com" title="Vedant Kumar <vsk@apple.com>"> <span class="fn">Vedant Kumar</span></a>
</span> changed
<a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED INVALID - [InstCombine][DebugInfo] replaceInstUsesWith() makes dbg.values point to mis-sized values"
href="https://bugs.llvm.org/show_bug.cgi?id=37883">bug 37883</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>INVALID
</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 INVALID - [InstCombine][DebugInfo] replaceInstUsesWith() makes dbg.values point to mis-sized values"
href="https://bugs.llvm.org/show_bug.cgi?id=37883#c2">Comment # 2</a>
on <a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED INVALID - [InstCombine][DebugInfo] replaceInstUsesWith() makes dbg.values point to mis-sized values"
href="https://bugs.llvm.org/show_bug.cgi?id=37883">bug 37883</a>
from <span class="vcard"><a class="email" href="mailto:vsk@apple.com" title="Vedant Kumar <vsk@apple.com>"> <span class="fn">Vedant Kumar</span></a>
</span></b>
<pre>Hold on, in this particular case, I think InstCombine is doing the right thing.
In the final output, !12 points to a 128-bit variable, which is consistent with
the value (<2 x i64> %and).
Could it be that you saw a mismatched size issue while working on
<a href="https://reviews.llvm.org/D48331">https://reviews.llvm.org/D48331</a> (see: <a href="https://paste.debian.net/1030112/">https://paste.debian.net/1030112/</a>)? In
the paste, I see:
call void @llvm.dbg.value(metadata <2 x i64> %and, metadata !11, metadata
!DIExpression()), !dbg !14
!10 = !DIBasicType(name: "ty64", size: 64, encoding: DW_ATE_unsigned)
!11 = !DILocalVariable(name: "2", scope: !6, file: !1, line: 2, type: !10)
This one does seem like a bona-fide bug, but it's not with ToT does.</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>