<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 - symbol lookup of runtime function with multiple definitions uses different definition in LTO vs. non-LTO"
href="https://bugs.llvm.org/show_bug.cgi?id=42273">42273</a>
</td>
</tr>
<tr>
<th>Summary</th>
<td>symbol lookup of runtime function with multiple definitions uses different definition in LTO vs. non-LTO
</td>
</tr>
<tr>
<th>Product</th>
<td>lld
</td>
</tr>
<tr>
<th>Version</th>
<td>unspecified
</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>ELF
</td>
</tr>
<tr>
<th>Assignee</th>
<td>unassignedbugs@nondot.org
</td>
</tr>
<tr>
<th>Reporter</th>
<td>efriedma@quicinc.com
</td>
</tr>
<tr>
<th>CC</th>
<td>llvm-bugs@lists.llvm.org, peter.smith@linaro.org
</td>
</tr></table>
<p>
<div>
<pre>Demonstration:
echo "struct S { int x[100]; }; void f(struct S* s) { s[0] = s[1]; }" > input.c
echo "void* memcpy(void* dest, const void* src, unsigned long size) {}" >
memcpy.c
clang -c input.c
clang -c memcpy.c
clang -flto input.c -c -o input-lto.o
ar cr memcpy.a memcpy.o
clang -shared memcpy.o -o memcpy.so
clang -shared -fuse-ld=lld input.o memcpy.a memcpy.so
nm a.out | grep memcpy
clang -shared -fuse-ld=lld input-lto.o memcpy.a memcpy.so
nm a.out | grep memcpy
The first "nm" (the non-LTO case) prints "T", the second (the LTO case) prints
"U".
Having multiple definitions of memcpy is sort of weird, but it would be nice to
behave consistently here.</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>