<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"><head><meta http-equiv=Content-Type content="text/html; charset=utf-8"><meta name=Generator content="Microsoft Word 15 (filtered medium)"><style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
span.EmailStyle18
{mso-style-type:personal-reply;
font-family:"Calibri",sans-serif;
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri",sans-serif;}
@page WordSection1
{size:612.0pt 792.0pt;
margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]--></head><body lang=EN-US link="#0563C1" vlink="#954F72"><div class=WordSection1><p class=MsoNormal>You are right that it<span style='font-family:"Times New Roman",serif'>’</span>s mostly a convenience for the front-ends. So they don<span style='font-family:"Times New Roman",serif'>’</span>t have to deal with boring things like padding and sizing things.<o:p></o:p></p><p class=MsoNormal>Otherwise it adds no semantic value. Object aliasing is not field sensitive in LLVM, so it doesn<span style='font-family:"Times New Roman",serif'>’</span>t matter. Though someone may want to add support for that in the future for languages where it<span style='font-family:"Times New Roman",serif'>’</span>s ok to do so.<o:p></o:p></p><p class=MsoNormal>FWIW, Alive2<span style='font-family:"Times New Roman",serif'>’</span>s GEP instruction works over bytes only (pairs of constant * %reg). Though I<span style='font-family:"Times New Roman",serif'>’</span>m not sure I would advocate to change LLVM<span style='font-family:"Times New Roman",serif'>’</span>s representation.<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal>Nuno<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><b>From:</b> Nikita Popov<br><b>Sent:</b> 13 July 2020 21:08<br><b>To:</b> llvm-dev <llvm-dev@lists.llvm.org><br><b>Subject:</b> [llvm-dev] Why are GEPs type based?<o:p></o:p></p><p class=MsoNormal><o:p> </o:p></p><div><div><p class=MsoNormal>Hi,<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>I've been wondering why LLVMs GEP instructions are based on types, rather than encoding the raw address calculation as a base pointer plus some scaled offsets (still in the form of a GEP, to retain provenance).<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>The type information does not seem particularly useful (shouldn't be used as an optimization base, because struct layouts lie), but increases the non-canonical IR space (there are many ways to encode the same GEP) and increases compile-time (optimizations need to constantly decompose GEPs, e.g. to get constant offsets).<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>What am I missing here?<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>Nikita,<o:p></o:p></p></div><div><p class=MsoNormal>Regards<o:p></o:p></p></div></div></div></body></html>