<html>
<head>
<base href="https://llvm.org/bugs/" />
</head>
<body><span class="vcard"><a class="email" href="mailto:bart.jacobs@cs.kuleuven.be" title="Bart Jacobs <bart.jacobs@cs.kuleuven.be>"> <span class="fn">Bart Jacobs</span></a>
</span> changed
<a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED INVALID - Info from .cfi directives should by default go into .debug_frame, not .eh_frame section"
href="https://llvm.org/bugs/show_bug.cgi?id=26191">bug 26191</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;">Status</td>
<td>NEW
</td>
<td>RESOLVED
</td>
</tr>
<tr>
<td style="text-align:right;">Resolution</td>
<td>---
</td>
<td>INVALID
</td>
</tr></table>
<p>
<div>
<b><a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED INVALID - Info from .cfi directives should by default go into .debug_frame, not .eh_frame section"
href="https://llvm.org/bugs/show_bug.cgi?id=26191#c7">Comment # 7</a>
on <a class="bz_bug_link
bz_status_RESOLVED bz_closed"
title="RESOLVED INVALID - Info from .cfi directives should by default go into .debug_frame, not .eh_frame section"
href="https://llvm.org/bugs/show_bug.cgi?id=26191">bug 26191</a>
from <span class="vcard"><a class="email" href="mailto:bart.jacobs@cs.kuleuven.be" title="Bart Jacobs <bart.jacobs@cs.kuleuven.be>"> <span class="fn">Bart Jacobs</span></a>
</span></b>
<pre>I think I figured out why people say the x86_64 ABI requires .eh_frame. The
main reason is: being able to get a correct backtrace, even in the absence of
debug symbols, is considered important for diagnosis and monitoring. Putting
the info in the .debug_frame section means it is only there in debug builds,
and it gets stripped when stripping debug info. That is considered undesirable.
I can appreciate that.
Still, I wish the use case of generating a backtrace were separated from the
use case of exception handling. In particular, I wish the Objective-C runtime
(through libunwind) did not interpret the presence of .eh_frame in my OCaml
binary as an invitation to walk into my OCaml stack frames (and crash on the
incorrect unwind info). I wish it were possible to mark frames as
"exception-safe" or not and to have the Objective-C runtime not walk into
frames not marked as exception-safe. But that's a separate request, I guess. So
I'm marking this as resolved.</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>