<html><head><meta http-equiv="Content-Type" content="text/html charset=utf-8"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class=""><br class=""><div><blockquote type="cite" class=""><div class="">On Dec 9, 2014, at 11:20 AM, Reid Kleckner <<a href="mailto:rnk@google.com" class="">rnk@google.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class="">How many atexit dtors do we have still? How long do you think this will stay?<br class=""></div></blockquote><div><br class=""></div>Unfortunately a lot. On the order of 1000. Most of those are cl::opt, so hopefully this won’t be needed for long.</div><div><br class=""><blockquote type="cite" class=""><div class=""><br class="">Given that we had such a struggle with the other symbol override in Signals.inc, I'm surprised you went this route. :)<br class=""></div></blockquote><div><br class=""></div><div>I really didn’t want to do this. Today WebKit builds their own dylib based on LLVM static archives, and we’re trying to get them to use something vended by LLVM. One of their requirements is no static destructors. This is how they are doing it today, and since the cl::opt stuff is still in a bit of limbo (<a href="http://reviews.llvm.org/D6207" class="">http://reviews.llvm.org/D6207</a>), I need a solution.</div><div><br class=""></div><div>I’m really hoping I can pull this out early 2015.</div><div><br class=""></div><div>-Chris</div><br class=""><blockquote type="cite" class=""><div class=""><br class="">REPOSITORY<br class=""> rL LLVM<br class=""><br class=""><a href="http://reviews.llvm.org/D6572" class="">http://reviews.llvm.org/D6572</a><br class=""><br class=""><br class=""></div></blockquote></div><br class=""></body></html>