<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 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;}
@font-face
{font-family:"MS PGothic";
panose-1:2 11 6 0 7 2 5 8 2 4;}
@font-face
{font-family:"\@MS PGothic";
panose-1:2 11 6 0 7 2 5 8 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0in;
margin-bottom:.0001pt;
font-size:12.0pt;
font-family:"MS PGothic","sans-serif";
mso-fareast-language:JA;}
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;}
p.MsoAcetate, li.MsoAcetate, div.MsoAcetate
{mso-style-priority:99;
mso-style-link:"Balloon Text Char";
margin:0in;
margin-bottom:.0001pt;
font-size:8.0pt;
font-family:"Tahoma","sans-serif";
mso-fareast-language:JA;}
span.EmailStyle17
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:#1F497D;}
span.BalloonTextChar
{mso-style-name:"Balloon Text Char";
mso-style-priority:99;
mso-style-link:"Balloon Text";
font-family:"Tahoma","sans-serif";
mso-fareast-language:JA;}
span.EmailStyle20
{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 1.0in 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"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">I actually would expect the LLVM JIT engine to generate the same code if you have everything prepped the same and use the same code generation options. If
you use MCJIT, in fact, it uses exactly the same code generation mechanism.<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Have you tried using MCJIT?<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Also, have you compared the IR being passed into the JIT to that being used in the clang and llc cases? If the IR is the same, I would expect the generated
code to be the same, unless there’s some additional optimization that isn’t being turned on in the JIT case.<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">-Andy<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"><o:p> </o:p></span></p>
<div>
<div style="border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0in 0in 0in">
<p class="MsoNormal"><b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> llvmdev-bounces@cs.uiuc.edu [mailto:llvmdev-bounces@cs.uiuc.edu]
<b>On Behalf Of </b>Micah Villmow<br>
<b>Sent:</b> Tuesday, August 20, 2013 9:06 AM<br>
<b>To:</b> ???; llvmdev@cs.uiuc.edu<br>
<b>Subject:</b> Re: [LLVMdev] Memory optimizations for LLVM JIT<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">I would not expect a JIT to produce as good of code as a static compiler. A JIT is supposed to run relatively fast, whereas a static compiler may take a lot
longer.<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Micah<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 #B5C4DF 1.0pt;padding:3.0pt 0in 0in 0in">
<p class="MsoNormal"><b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">
<a href="mailto:llvmdev-bounces@cs.uiuc.edu">llvmdev-bounces@cs.uiuc.edu</a> [<a href="mailto:llvmdev-bounces@cs.uiuc.edu">mailto:llvmdev-bounces@cs.uiuc.edu</a>]
<b>On Behalf Of </b>???<br>
<b>Sent:</b> Tuesday, August 20, 2013 12:23 AM<br>
<b>To:</b> <a href="mailto:llvmdev@cs.uiuc.edu">llvmdev@cs.uiuc.edu</a><br>
<b>Subject:</b> [LLVMdev] Memory optimizations for LLVM JIT<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">Hello. </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">I'm new to LLVM and I've got some problems with LLVM JIT.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">I have set ExecutionEngine as CodeGenOpt::Aggressive and PassManagerBuilder.OptLevel as 3. (mem2reg and GVN are included)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">However, the machine code generated by JIT is not as good as that generated by clang or llc.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">Here is an example:</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">--------------------------------------------------------------------</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> source fragment ==> clang or llc</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">struct {</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> uint64_t a[10];</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">} *p;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> mov 0x8(%rax),%rdx</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">p->a[2] = p->a[1]; mov %rdx,0x10(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">p->a[3] = p->a[1]; ==> mov %rdx,0x18(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">p->a[4] = p->a[2]; mov %rdx,0x20(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">p->a[5] = p->a[4]; mov %rdx,0x28(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">--------------------------------------------------------------------</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> JIT (map p to GlobalVariable) ==> JIT (map p to constant GlobalVariable)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 1* movabsq $0x18c6b88, %rax 1* movabsq $0x18c6b88, %rax</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 2* movq (%rax), %rcx // p 2* movq (%rax), %rax</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 3* movq 0x8(%rcx), %rdx // a[1] 3* movq 0x8(%rax), %rcx</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 4* movq %rdx, 0x10(%rcx) // a[2] 4* movq %rcx, 0x10(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 5 movq (%rax), %rcx 5</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 6 movq 0x8(%rcx), %rdx 6 movq 0x8(%rax), %rcx</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 7* movq %rdx, 0x18(%rcx) ==> 7* movq %rcx, 0x18(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 8 movq (%rax), %rcx 8</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New""> 9 movq 0x10(%rcx), %rdx 9 movq 0x10(%rax), %rcx</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">10* movq %rdx, 0x20(%rcx) 10* movq %rcx, 0x20(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">11 movq (%rax), %rax 11</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">12 movq 0x20(%rax), %rcx 12</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">13* movq %rcx, 0x28(%rax) 13* movq %rcx, 0x28(%rax)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">----------------------------------------------------------------------</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">A GlobalValue was declared and mapped to the variable p.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">Some LLVM IR instructions were created according to those generated by LLVM from source.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">I.e., load p, load a[1] based on p, load p again, store a[2] based on p, etc.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">The machine code turned out to be slightly optmized, as shown on the left.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">Things were getting better after the GlobalVariable of p was set as a constant.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">Redundant Loads of p (line 5, 8 and 11) were removed, and so was line 12 because of line 10.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">However, I could not make it better any more, although optimal machine code just need those marked with '*'.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">It seems that store instructions block the optimizations across them.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">I.e., line 3&6 or 4&9 are similar to line 10&12, but they are not optimized.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">The store (line 4 or 7) between them obviously has no alias problem.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">My question is: how to make LLVM JIT optimize this code? </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">Did I miss anything, or need I write some kind of optimization pass?</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-family:"Courier New"">I will be grateful for any help you can provide.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
</div>
</div>
</div>
</body>
</html>