<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=koi8-r">
<meta name="Generator" content="Microsoft Word 14 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
@font-face
{font-family:Tahoma;
panose-1:2 11 6 4 3 5 4 4 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
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;}
p.MsoAcetate, li.MsoAcetate, div.MsoAcetate
{mso-style-priority:99;
mso-style-link:"Balloon Text Char";
margin:0cm;
margin-bottom:.0001pt;
font-size:8.0pt;
font-family:"Tahoma","sans-serif";}
p.MsoListParagraph, li.MsoListParagraph, div.MsoListParagraph
{mso-style-priority:34;
margin-top:0cm;
margin-right:0cm;
margin-bottom:0cm;
margin-left:36.0pt;
margin-bottom:.0001pt;
font-size:11.0pt;
font-family:"Calibri","sans-serif";}
span.BalloonTextChar
{mso-style-name:"Balloon Text Char";
mso-style-priority:99;
mso-style-link:"Balloon Text";
font-family:"Tahoma","sans-serif";}
p.msonormal0, li.msonormal0, div.msonormal0
{mso-style-name:msonormal;
mso-margin-top-alt:auto;
margin-right:0cm;
mso-margin-bottom-alt:auto;
margin-left:0cm;
font-size:12.0pt;
font-family:"Times New Roman","serif";}
span.EmailStyle20
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:windowtext;}
span.EmailStyle21
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:#1F497D;}
span.EmailStyle22
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:#1F497D;}
span.EmailStyle23
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:#1F497D;}
span.EmailStyle24
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:#1F497D;}
span.EmailStyle25
{mso-style-type:personal-reply;
font-family:"Calibri","sans-serif";
color:#1F497D;}
.MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;}
@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-GB" link="#0563C1" vlink="#954F72">
<div class="WordSection1">
<p class="MsoNormal"><span style="color:#1F497D">I feel that both solutions offer similar functionality with slightly different APIs. Just my general feeling is that these are rather properties of an application and have nothing to do with target specific details.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">></span><span style="color:#1F497D">
<span lang="EN-US">The only problem I see here if someone else would want to use function metadata attachment for some other reason, so we will need to search and/or merge it.<o:p></o:p></span></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">@Stas, would it not be the same if other target-dependent attributes</span><span style="color:#1F497D"> are attached too?<o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">> </span><span lang="EN-US" style="color:#1F497D">The function metadata also requires a commonly agreed attribute name. In this aspect it is similar to target-dependent attribute-value pair. However it may be
more difficult to manipulate.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">@Sam, could you provide more details where attributes would be more beneficial to use?<o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">Thanks,<o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">Anastasia<o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"><o:p> </o:p></span></p>
<div>
<div style="border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> Mekhanoshin, Stanislav [mailto:Stanislav.Mekhanoshin@amd.com]
<br>
<b>Sent:</b> 11 May 2016 19:39<br>
<b>To:</b> Liu, Yaxun (Sam); Anastasia Stulova; cfe-dev (cfe-dev@lists.llvm.org); Pan, Xiuli; Bader, Alexey (alexey.bader@intel.com)<br>
<b>Cc:</b> Stellard, Thomas; Sumner, Brian; nd<br>
<b>Subject:</b> RE: [OpenCL] Representing kernel attributes by LLVM target-dependent attribute-value pairs instead of metadata<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Sam,</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">I guess metadata attached right to the function should work as well. Nothing would hold us from forming the same list of strings in a single metadata entry. The only problem I see here if someone
else would want to use function metadata attachment for some other reason, so we will need to search and/or merge it.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Given that consideration attributes looks better to me, they are already a top-level list.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<div>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Stas</span><span lang="EN-US"><o:p></o:p></span></p>
</div>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<div>
<div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US">From:</span></b><span lang="EN-US"> Liu, Yaxun (Sam)
<br>
<b>Sent:</b> 11 ΝΑΡ 2016 Η. 10:18<br>
<b>To:</b> Anastasia Stulova <<a href="mailto:Anastasia.Stulova@arm.com">Anastasia.Stulova@arm.com</a>>; Mekhanoshin, Stanislav <<a href="mailto:Stanislav.Mekhanoshin@amd.com">Stanislav.Mekhanoshin@amd.com</a>>; cfe-dev (<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>)
<<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>>; Pan, Xiuli <<a href="mailto:xiuli.pan@intel.com">xiuli.pan@intel.com</a>>; Bader, Alexey (<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>) <<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>><br>
<b>Cc:</b> Stellard, Thomas <<a href="mailto:Tom.Stellard@amd.com">Tom.Stellard@amd.com</a>>; Sumner, Brian <<a href="mailto:Brian.Sumner@amd.com">Brian.Sumner@amd.com</a>>; nd <<a href="mailto:nd@arm.com">nd@arm.com</a>><br>
<b>Subject:</b> RE: [OpenCL] Representing kernel attributes by LLVM target-dependent attribute-value pairs instead of metadata<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Hi Anastasia,</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Target-dependent attribute-value pair is not limited to one target. It is just a pair of strings attached to a function. It can be used by different targets as far as they accept a commonly agreed
attribute name.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">The function metadata also requires a commonly agreed attribute name. In this aspect it is similar to target-dependent attribute-value pair. However it may be more difficult to manipulate.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Stas, any comments on this?</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Thanks.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Sam</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<div>
<div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US">From:</span></b><span lang="EN-US"> Anastasia Stulova [<a href="mailto:Anastasia.Stulova@arm.com">mailto:Anastasia.Stulova@arm.com</a>]
<br>
<b>Sent:</b> Wednesday, May 11, 2016 12:32 PM<br>
<b>To:</b> Mekhanoshin, Stanislav <<a href="mailto:Stanislav.Mekhanoshin@amd.com">Stanislav.Mekhanoshin@amd.com</a>>; Liu, Yaxun (Sam) <<a href="mailto:Yaxun.Liu@amd.com">Yaxun.Liu@amd.com</a>>; cfe-dev (<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>)
<<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>>; Pan, Xiuli <<a href="mailto:xiuli.pan@intel.com">xiuli.pan@intel.com</a>>; Bader, Alexey (<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>) <<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>><br>
<b>Cc:</b> Stellard, Thomas <<a href="mailto:Tom.Stellard@amd.com">Tom.Stellard@amd.com</a>>; Sumner, Brian <<a href="mailto:Brian.Sumner@amd.com">Brian.Sumner@amd.com</a>>; nd <<a href="mailto:nd@arm.com">nd@arm.com</a>><br>
<b>Subject:</b> RE: [OpenCL] Representing kernel attributes by LLVM target-dependent attribute-value pairs instead of metadata<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">Hi Sam,</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">Interesting idea, I am just concerned whether target-dependent attribute-value pairs are really correct for this use case. Because kernel attributes aren’t really target dependent.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">However, I agree that current approach is painful enough. Do you think switching to function metadata (see llvm@r235783) would help instead? This way we will also have the kernel attributes directly accessible
from the function, simplifying querying them (as compared to what we have at the moment).</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">Thanks,</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D">Anastasia</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<div>
<div style="border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span lang="EN-US" style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> Mekhanoshin, Stanislav [<a href="mailto:Stanislav.Mekhanoshin@amd.com">mailto:Stanislav.Mekhanoshin@amd.com</a>]
<br>
<b>Sent:</b> 09 May 2016 21:14<br>
<b>To:</b> Liu, Yaxun (Sam); cfe-dev (<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>); Anastasia Stulova; Pan, Xiuli; Bader, Alexey (<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>)<br>
<b>Cc:</b> Stellard, Thomas; Sumner, Brian<br>
<b>Subject:</b> RE: [OpenCL] Representing kernel attributes by LLVM target-dependent attribute-value pairs instead of metadata</span><span lang="EN-US"><o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"> <span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">It will make our life much easier I would say.</span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<div>
<p class="MsoNormal"><span lang="EN-US" style="color:#1F497D">Stas</span><span lang="EN-US"><o:p></o:p></span></p>
</div>
<p class="MsoNormal"><span lang="RU" style="color:#1F497D"> </span><span lang="EN-US"><o:p></o:p></span></p>
<div>
<div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US">From:</span></b><span lang="EN-US"> Liu, Yaxun (Sam)
<br>
<b>Sent:</b> Monday, May 09, 2016 1:05 PM<br>
<b>To:</b> cfe-dev (<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>) <<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a>>; 'anastasia.stulova@arm.com' <<a href="mailto:anastasia.stulova@arm.com">anastasia.stulova@arm.com</a>>;
Pan, Xiuli <<a href="mailto:xiuli.pan@intel.com">xiuli.pan@intel.com</a>>; Bader, Alexey (<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>) <<a href="mailto:alexey.bader@intel.com">alexey.bader@intel.com</a>><br>
<b>Cc:</b> Stellard, Thomas <<a href="mailto:Tom.Stellard@amd.com">Tom.Stellard@amd.com</a>>; Sumner, Brian <<a href="mailto:Brian.Sumner@amd.com">Brian.Sumner@amd.com</a>>; Mekhanoshin, Stanislav <<a href="mailto:Stanislav.Mekhanoshin@amd.com">Stanislav.Mekhanoshin@amd.com</a>><br>
<b>Subject:</b> [OpenCL] Representing kernel attributes by LLVM target-dependent attribute-value pairs instead of metadata<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><span lang="RU"> </span><span lang="EN-US"><o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">Hi,<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">When SPIR 1.2/2.0 spec was developed, there was no target-dependent attribute-value pair in LLVM. As such, kernel attributes were represented as metadata.<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">This caused lots of inconvenience since these metadata are difficult to manipulate/change when kernel functions went through transformations in backends.<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">Now LLVM supports target-dependent attribute-value pair, I am wondering whether it is time to use that to represent kernel attributes.<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">e.g. something like (just to give an idea, not exact llvm IR)<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">kernel void foo(global int*a, sampler_t s); #1<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">#1 = {<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">reqd_work_group_size=”1 1 1”,<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">kernel_arg_type=”int *, sampler_t”,<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">…<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">}<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">basically we keep the info conveyed by the original metadata but attach them to the kernel function as attribute/value pair. This will make these attributes much easier to manipulate/change.<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">Any feedbacks? Thanks.<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US"> <o:p></o:p></span></p>
<p class="MsoNormal"><span lang="EN-US">Sam<o:p></o:p></span></p>
</div>
</body>
</html>