<html><head><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></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 Nov 18, 2016, at 9:20 AM, David Li <<a href="mailto:davidxl@google.com" class="">davidxl@google.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><div class="">davidxl added inline comments.<br class=""><br class=""><br class="">================<br class="">Comment at: include/llvm/Analysis/InlineCost.h:39<br class=""> /// Use when -O3 is specified.<br class="">-const int OptAggressiveThreshold = 275;<br class="">+const int OptAggressiveThreshold = 250;<br class=""><br class="">----------------<br class="">Any reason to reduce O3 threshold?<br class=""></div></div></blockquote><div><br class=""></div>Look at the compile time jumps of > 10% at <a href="http://lab.llvm.org:8080/green/view/Compile Time/" class="">http://lab.llvm.org:8080/green/view/Compile%20Time/</a></div><div><br class=""></div><div>When there is no significant performance benefit (measured on wide range of standard benchmark and apps) reducing the threshold is the sensitive approach. </div><div><br class=""></div><div>-Gerolf</div><div><br class=""><blockquote type="cite" class=""><div class=""><div class=""><br class="">Repository:<br class=""> rL LLVM<br class=""><br class=""><a href="https://reviews.llvm.org/D26848" class="">https://reviews.llvm.org/D26848</a><br class=""><br class=""><br class=""><br class=""></div></div></blockquote></div><br class=""></body></html>