<div><div dir="auto">Hi Siddzharth,</div><div dir="auto"><br></div><div dir="auto">Is it possible to implement equality saturation directly on (a subset of) LLVM IR, such that you don't need to create yet-another compiler IR?</div><div dir="auto"><br></div><div dir="auto">Thanks</div><div dir="auto">Hongbin </div><br><div class="gmail_quote"><div>On Sun, Sep 17, 2017 at 2:53 AM (IIIT) Siddharth Bhat via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org">llvm-dev@lists.llvm.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div>I've started hacking up a branch with a rough implementation of equality saturation. This needs me to add objects such as <font face="monospace">PEGBasicBlock</font>, <font face="monospace">PEGInstruction</font>, <font face="monospace">PEGTerminatorInstruction</font>, etc. (PEG stands for "program expression block"). I wish to share most of the infrastructure that exists - I believe this is already done with <font face="monospace">Machine*</font>. <div><br></div><div>How do I do this for PEGs? I'd also like to get analysis passes for "free" - particularly, dominance and loop analyses. </div><div><br></div><div>Thanks,</div><div>Siddzharth</div></div><div>-- <br></div><div class="m_-8041150019620572716gmail_signature" data-smartmail="gmail_signature"><div>Sending this from my phone, please excuse any typos!</div></div>
_______________________________________________<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/mailman/listinfo/llvm-dev</a><br>
</blockquote></div></div>