<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Jul 24, 2017 at 2:29 PM, John Regehr via llvm-dev <span dir="ltr"><<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="gmail-">On 7/24/17 3:16 PM, Hongbin Zheng via llvm-dev wrote:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
Thanks, maybe we could use ScalarEvolution in LazyValueInfo if it is<br>
available?<br>
</blockquote>
<br></span>
This should be fairly easy to try, if you want to propose a patch and run some experiments. The question is whether the benefit in terms of improved optimization power is worth the compile-time cost.</blockquote><div> Ok</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"> Also, the last time I looked, there seemed to be a fair amount of room for improvement in LVI and ConstantRange.</blockquote><div>Could you point some out?</div><div><br></div><div><br></div><div>Thanks</div><div>Hongbin</div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="gmail-HOEnZb"><font color="#888888"><br>
<br>
John</font></span><div class="gmail-HOEnZb"><div class="gmail-h5"><br>
<br>
______________________________<wbr>_________________<br>
LLVM Developers mailing list<br>
<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a><br>
<a href="http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev" rel="noreferrer" target="_blank">http://lists.llvm.org/cgi-bin/<wbr>mailman/listinfo/llvm-dev</a><br>
</div></div></blockquote></div><br></div></div>