<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 12 (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:0in;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri","sans-serif";}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
span.EmailStyle17
{mso-style-type:personal-compose;
font-family:"Calibri","sans-serif";
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;}
@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="blue" vlink="purple">
<div class="WordSection1">
<p class="MsoNormal">Since Clang does not seem to allow type casts, such as uchar4 to float4, between vector types, it seems it is necessary to write them as element by element conversions, such as<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New";color:blue">typedef</span><span style="font-size:10.0pt;font-family:"Courier New"">
<span style="color:blue">float</span> float4 __attribute__((ext_vector_type(4)));<o:p></o:p></span></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New";color:blue">typedef</span><span style="font-size:10.0pt;font-family:"Courier New"">
<span style="color:blue">unsigned</span> <span style="color:blue">char</span> uchar4 __attribute__((ext_vector_type(4)));<o:p></o:p></span></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New""><o:p> </o:p></span></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New"">float4 to_float4(uchar4 in)<o:p></o:p></span></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New"">{<o:p></o:p></span></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New""> float4 out = {in.x, in.y, in.z, in.w};<o:p></o:p></span></p>
<p class="MsoNormal" style="text-autospace:none"><span style="font-size:10.0pt;font-family:"Courier New"">
<span style="color:blue">return</span> out;<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New"">}<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""><o:p> </o:p></span></p>
<p class="MsoNormal">Running this code through “clang –c –emit-llvm” and then through “opt –O2 –S”, produces the following IR:<o:p></o:p></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New"">define <4 x float> @to_float4(i32 %in.coerce) nounwind uwtable readnone {<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New"">entry:<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %0 = bitcast i32 %in.coerce to <4 x i8><o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %1 = extractelement <4 x i8> %0, i32 0<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %conv = uitofp i8 %1 to float<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %vecinit = insertelement <4 x float> undef, float %conv, i32 0<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %2 = extractelement <4 x i8> %0, i32 1<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %conv2 = uitofp i8 %2 to float<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %vecinit3 = insertelement <4 x float> %vecinit, float %conv2, i32 1<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %3 = extractelement <4 x i8> %0, i32 2<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %conv4 = uitofp i8 %3 to float<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %vecinit5 = insertelement <4 x float> %vecinit3, float %conv4, i32 2<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %4 = extractelement <4 x i8> %0, i32 3<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %conv6 = uitofp i8 %4 to float<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %vecinit7 = insertelement <4 x float> %vecinit5, float %conv6, i32 3<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> ret <4 x float> %vecinit7<o:p></o:p></span></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">Which does the cast as a sequence of scalar operations, whereas it could be done as<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> %1 = uitofp <4 x i8> %0 to <4 x float><o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Courier New""> ret <4 x float> %1<o:p></o:p></span></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">It seemed to me that the recently committed basic block vectorizer might be able to do this kind of optimization, but the current version does not do so.<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">Is this optimization the kind of thing that the bb-vectorizer is intended to be able to do? And, if so, do you have any suggestions as to how that may be done? Or, if not, can you suggest another possible way to parallelize this kind of
code?<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">Thanks,<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">Preston<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">--<o:p></o:p></p>
<p class="MsoNormal">Preston Gurd <preston.gurd@intel.com><o:p></o:p></p>
<p class="MsoNormal"> Intel Waterloo<o:p></o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
</body>
</html>