<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)">
<!--[if !mso]><style>v\:* {behavior:url(#default#VML);}
o\:* {behavior:url(#default#VML);}
w\:* {behavior:url(#default#VML);}
.shape {behavior:url(#default#VML);}
</style><![endif]--><style><!--
/* Font Definitions */
@font-face
{font-family:宋体;
panose-1:2 1 6 0 3 1 1 1 1 1;}
@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;}
@font-face
{font-family:"\@宋体";
panose-1:2 1 6 0 3 1 1 1 1 1;}
@font-face
{font-family:Tahoma;
panose-1:2 11 6 4 3 5 4 4 2 4;}
@font-face
{font-family:Verdana;
panose-1:2 11 6 4 3 5 4 4 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
margin-bottom:.0001pt;
font-size:12.0pt;
font-family:"Times New Roman",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-reply;
font-family:"Calibri",sans-serif;
color:#1F497D;}
.MsoChpDefault
{mso-style-type:export-only;
font-size:10.0pt;}
@page WordSection1
{size:8.5in 11.0in;
margin:1.0in 1.25in 1.0in 1.25in;}
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"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
<p class="MsoNormal"><a name="_MailEndCompose"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">The load store optimizer did not do aggressive merge.<o:p></o:p></span></a></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">Normally the successive load instructions are not too far.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">The performance difference is much higher than I thought.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">So the performance number comes for SKL platform? Have you tried this patch on a BDW?<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">The performance behavior you observed may not be applied to other platform.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D">How much will the patch affect the compile time?<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri",sans-serif;color:#1F497D"><o:p> </o:p></span></p>
<div style="border:none;border-left:solid blue 1.5pt;padding:0in 0in 0in 4.0pt">
<div>
<div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0in 0in 0in">
<p class="MsoNormal"><b><span style="font-size:11.0pt;font-family:"Calibri",sans-serif">From:</span></b><span style="font-size:11.0pt;font-family:"Calibri",sans-serif"> yan.wang [mailto:yan.wang@linux.intel.com]
<br>
<b>Sent:</b> Friday, March 10, 2017 10:52 AM<br>
<b>To:</b> Song, Ruiling <ruiling.song@intel.com>; beignet <beignet@lists.freedesktop.org><br>
<b>Subject:</b> Re: Re: [Beignet] [PATCH v2] Provide more possible candidate of load/store as possible.<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">It comes from darktable perforamnce tuning.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">For float type, maxVecSize is 4, so maxLimit = 4 * 8 = 32.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">I am not sure the reason of maxLimit = maxVecSize * 8.</span><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:#1F497D"><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">32 is too samll for saerching and could not find more available load after leading load.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">It will improve eaw_decompose kernel of darktable from 2.1876s to 1.8855s because reduce send from 3 send (2 float, 2 float, 1 float) to 2 send (4 float, 1 float).<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">There is another issue when compiing eaw_decompose kernel and I will submit another patch for it.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">At least need set one low bound for maxLimit like 150 to avoid seaching range too slow.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"><o:p> </o:p></span></p>
</div>
<div class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">
<hr size="1" width="263" style="width:157.5pt" noshade="" style="color:#B5C4DF" align="left">
</span></div>
<div>
<div style="margin-left:7.5pt;margin-top:7.5pt;margin-right:7.5pt;margin-bottom:7.5pt">
<div>
<p class="MsoNormal"><span style="font-size:10.0pt;font-family:"Verdana",sans-serif;color:black">yan.wang<o:p></o:p></span></p>
</div>
</div>
</div>
<blockquote style="margin-left:6.0pt">
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"> <o:p></o:p></span></p>
</div>
<div style="border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0in 0in 0in">
<div>
<div>
<p class="MsoNormal" style="background:#EFEFEF"><b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black">From:</span></b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black"> <a href="mailto:ruiling.song@intel.com">Song,
Ruiling</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal" style="background:#EFEFEF"><b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black">Date:</span></b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black"> 2017-03-10 10:39<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal" style="background:#EFEFEF"><b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black">To:</span></b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black"> <a href="mailto:yan.wang@linux.intel.com">yan.wang@linux.intel.com</a>;
<a href="mailto:beignet@lists.freedesktop.org">beignet@lists.freedesktop.org</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal" style="background:#EFEFEF"><b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black">Subject:</span></b><span style="font-size:9.0pt;font-family:"Tahoma",sans-serif;color:black"> Re: [Beignet] [PATCH v2] Provide more
possible candidate of load/store as possible.<o:p></o:p></span></p>
</div>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"> <o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"> <o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> -----Original Message-----<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> From: Beignet [<a href="mailto:beignet-bounces@lists.freedesktop.org">mailto:beignet-bounces@lists.freedesktop.org</a>] On Behalf Of<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<a href="mailto:yan.wang@linux.intel.com">yan.wang@linux.intel.com</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> Sent: Thursday, March 9, 2017 5:41 PM<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> To:
<a href="mailto:beignet@lists.freedesktop.org">beignet@lists.freedesktop.org</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> Cc: Yan Wang <<a href="mailto:yan.wang@linux.intel.com">yan.wang@linux.intel.com</a>><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> Subject: [Beignet] [PATCH v2] Provide more possible candidate of load/store as<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> possible.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> From: Yan Wang <<a href="mailto:yan.wang@linux.intel.com">yan.wang@linux.intel.com</a>><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> Avoid searching range too small in some case like vector of float.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> It will lead more load/store merged for improving perforamnce.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> Signed-off-by: Yan Wang <<a href="mailto:yan.wang@linux.intel.com">yan.wang@linux.intel.com</a>><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> ---<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> backend/src/llvm/llvm_loadstore_optimization.cpp | 2 +-<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> 1 file changed, 1 insertion(+), 1 deletion(-)<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> diff --git a/backend/src/llvm/llvm_loadstore_optimization.cpp<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> b/backend/src/llvm/llvm_loadstore_optimization.cpp<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> index e797e98..e569a8e 100644<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> --- a/backend/src/llvm/llvm_loadstore_optimization.cpp<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> +++ b/backend/src/llvm/llvm_loadstore_optimization.cpp<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> @@ -180,7 +180,7 @@ namespace gbe {<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> BasicBlock::iterator J = start;<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> ++J;<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> - unsigned maxLimit = maxVecSize * 8;<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> + unsigned maxLimit = std::max(maxVecSize * 8, 150u);<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"> <o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">Could you give some performance number against some known benchmarks?<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">Please select some complex enough OpenCL kernel. Maybe luxmark? Darktable?<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">How it would benefit the runtime performance and how much it would hurt the compile-time performance?<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">So we could know whether the change is reasonable.<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"> <o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">Thanks!<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">Ruiling<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> bool reordered = false;<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> for(unsigned ss = 0; J != E && ss <= maxLimit; ++ss, ++J) {<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> --<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> 2.7.4<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> _______________________________________________<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">> Beignet mailing list<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<a href="mailto:Beignet@lists.freedesktop.org">Beignet@lists.freedesktop.org</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">>
<a href="https://lists.freedesktop.org/mailman/listinfo/beignet">https://lists.freedesktop.org/mailman/listinfo/beignet</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">_______________________________________________<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black">Beignet mailing list<o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"><a href="mailto:Beignet@lists.freedesktop.org">Beignet@lists.freedesktop.org</a><o:p></o:p></span></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:10.5pt;font-family:"Segoe UI",sans-serif;color:black"><a href="https://lists.freedesktop.org/mailman/listinfo/beignet">https://lists.freedesktop.org/mailman/listinfo/beignet</a><o:p></o:p></span></p>
</div>
</div>
</blockquote>
</div>
</div>
</body>
</html>