<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=us-ascii">
<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;}
@font-face
{font-family:"Segoe UI";
panose-1:2 11 5 2 4 2 4 2 2 3;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:#0563C1;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:#954F72;
text-decoration:underline;}
span.EmailStyle17
{mso-style-type:personal-compose;
font-family:"Calibri",sans-serif;
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;
font-family:"Calibri",sans-serif;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.0in 1.0in 1.0in;}
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">Hi!<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">I would appreciate some feedback from someone with experience in SCEV/SE. D39346 tries to fix an issue in LV (<span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">PR34965)</span> that exposes a limitation in SCEV/SE. The best
solution to the LV issue might not be a fix at SCEV/SE level but we may want to report/address SCEV/SE limitation as well.
<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">For the snippet below, LV expects SE to return a SCEVAddRecExpr for
</span>%21. However, SE returns <span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">
((4 * (zext i32 (2 + %18) to i64))<nuw><nsw> + undef)<nsw>, probably because it’s not able to deduce that</span> %bc.resume.val1 = %bc.resume.val + 1.
<span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">You can find further information in
</span>D39346 and the full test in <span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">
PR34965 (</span>“IR_after_createVectorizedLoopSkeleton” attachment)<span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif"><o:p> </o:p></span></p>
<p class="MsoNormal">I could open a new bug against SCEV/SE if a fix or extension is feasible at that level.<o:p></o:p></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">Thanks,<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Segoe UI",sans-serif">Diego</span><o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">----<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">.outer: ; preds = %2, %0<o:p></o:p></p>
<p class="MsoNormal"> …<o:p></o:p></p>
<p class="MsoNormal"> %.ph2 = phi i32 [ 62, %2 ], [ 110, %0 ] <o:p></o:p></p>
<p class="MsoNormal"> %5 = add i32 %.ph2, 1<o:p></o:p></p>
<p class="MsoNormal"> …<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">vector.ph: ; preds = %.outer<o:p></o:p></p>
<p class="MsoNormal"> …<o:p></o:p></p>
<p class="MsoNormal"> %ind.end = add i32 %5, %n.vec<o:p></o:p></p>
<p class="MsoNormal"> %ind.end2 = add i32 %.ph2, %n.vec<o:p></o:p></p>
<p class="MsoNormal"> …<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">middle.block: ; preds = %vector.body.latch<o:p></o:p></p>
<p class="MsoNormal"> …<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">scalar.ph: ; preds = %middle.block, %.outer<o:p></o:p></p>
<p class="MsoNormal"> %bc.resume.val = phi i32 [ %ind.end, %middle.block ], [ %5, %.outer ]<o:p></o:p></p>
<p class="MsoNormal"> %bc.resume.val1 = phi i32 [ %ind.end2, %middle.block ], [ %.ph2, %.outer ]<o:p></o:p></p>
<p class="MsoNormal"> br label %16<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">; <label>:16: ; preds = %16, %scalar.ph<o:p></o:p></p>
<p class="MsoNormal"> %17 = phi i32 [ %bc.resume.val, %scalar.ph ], [ %23, %16 ]<o:p></o:p></p>
<p class="MsoNormal"> %18 = phi i32 [ %bc.resume.val1, %scalar.ph ], [ %17, %16 ]<o:p></o:p></p>
<p class="MsoNormal"> %19 = add i32 %18, 2<o:p></o:p></p>
<p class="MsoNormal"> %20 = zext i32 %19 to i64<o:p></o:p></p>
<p class="MsoNormal"> %21 = getelementptr inbounds i32, i32 addrspace(1)* undef, i64 %20<o:p></o:p></p>
<p class="MsoNormal"> %22 = ashr i32 undef, %4<o:p></o:p></p>
<p class="MsoNormal"> store i32 %22, i32 addrspace(1)* %21, align 4<o:p></o:p></p>
<p class="MsoNormal"> %23 = add i32 %17, 1<o:p></o:p></p>
<p class="MsoNormal"> %24 = icmp sgt i32 %23, 61<o:p></o:p></p>
<p class="MsoNormal"> br i1 %24, label %._crit_edge.loopexit.preheader, label %16<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
</body>
</html>