<div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr">Dear Bhatu and Mehdi,<div><br></div><div>Thank you for your helpful suggestions. </div><div><br></div><div>Indeed -print-after-all option is quite useful to find out the differences between the two approaches. As Bhatu pointed out, when providing -O3 as an option some of the passes are initialized differently when compared to simply passing a list of passes to opt. Some of the passes even depend on OptLevel as a parameter which influences the behavior of a pass, e.g. the following passes directly depend on the OptLevel parameter:</div><div>* createSimpleLoopUnrollPass()</div><div>* createLoopUnrollAndJamPass()</div><div>* createLoopUnrollPass()</div><div><br></div><div>However, the only way of setting OptLevel and SizeLevel parameters through CLI is by providing -O<> optimization flag, which in turn results in all of the associated passes being added to the list of passes. Is there a way of setting OptLevel = 3 and SizeLevel = 0 without running all the passes associated with -O3? That would probably get me closest to replicating O3, but I cannot figure out how to do it.</div><div><br clear="all"><div><div dir="ltr" class="gmail_signature"><div dir="ltr"><div><div dir="ltr">Kind Regards,<div><br></div><div>Rahim Mammadli</div></div></div></div></div></div><br></div></div></div></div></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, 13 May 2019 at 09:49, Bhatu <<a href="mailto:CS12B1010@iith.ac.in">CS12B1010@iith.ac.in</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr">I think this has to do with how the pass manager is populated when we give -O3 vs when we give particular pass names.<div>Some passes have multiple createXYZPass() methods that accept arguments too. These methods call non-default pass constructors, which in turn cause the passes to behave in a different manner.</div><div>eg:</div><div><div>Pass *llvm::createLICMPass() { return new LegacyLICMPass(); }</div><div>Pass *llvm::createLICMPass(unsigned LicmMssaOptCap,</div><div> unsigned LicmMssaNoAccForPromotionCap) {</div><div> return new LegacyLICMPass(LicmMssaOptCap, LicmMssaNoAccForPromotionCap);</div><div>}</div></div><div>or</div><div><div>Pass *createLoopVectorizePass() { return new LoopVectorize(); }</div><div>Pass *createLoopVectorizePass(bool InterleaveOnlyWhenForced,</div><div> bool VectorizeOnlyWhenForced) {</div><div> return new LoopVectorize(InterleaveOnlyWhenForced, VectorizeOnlyWhenForced);</div><div>}</div></div><div>When we give pass names, opt calls the default constructor (eg: LoopVectorize()) whereas when we give O3, it can call a different version.</div><div>You can check in PassManagerBuilder.cpp (populateModulePassManager, populateFunctionPassManager) to see where different versions are being populated. Those must be the points in the pipeline where the IR starts differing. </div></div></div></div></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Sat, May 11, 2019 at 10:09 PM Mehdi AMINI via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div dir="ltr">Hi,</div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Thu, May 9, 2019 at 5:20 PM Rahim Mammadli via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr">Dear developers,<div><br></div><div>I am trying to reproduce the results of applying opt -O3 to a source file in the form of LLVM IR. I want to get the same IR by manually ordering the passes used by O3 and passing them to opt. </div><div><br></div><div>To illustrate what I am doing on an example, as an input I use linpack benchmark from the LLVM test suite[1]:</div><div><br></div><div>1. First I produce the intermediate representation using clang:</div><div>clang -O3 -Xclang -disable-llvm-optzns -emit-llvm -S linpack-pc.c -o linpack-pc.ll<br></div><div><br></div><div>2. Then I use opt to optimize the IR:</div><div>opt -S -O3 -o linpack-pc-3.ll linpack-pc.ll<br></div><div><br></div><div>Now my goal is to produce the IR identical to linpack-pc-3.ll by passing a sequence of optimizations to opt. To get the list of optimizations used by opt for O3, I run this:</div><div>opt -O3 -disable-output -debug-pass=Arguments linpack-pc.ll<br></div><div><br></div><div>Which produces (shortened to avoid wasting space):</div><div><div>Pass Arguments: -tti -targetlibinfo -tbaa ...</div><div>Pass Arguments: -targetlibinfo -tti -tbaa ...</div><div>Pass Arguments: -domtree</div><div><br></div></div><div>So apparently there are three sequences of passes applied to IR as part of O3. I wasn't able to reproduce the same IR as linpack-pc-3.ll using these passes, I tried applying passes sequentially or concatenating them and passing as a single sequence to opt. Neither produced the needed output. Moreover the performance of the final executable downgraded by about 35%. I'm using LLVM 3.8 and my OS is Ubuntu 16.04.</div><div><br></div><div>[1] <a href="https://github.com/llvm/llvm-test-suite/blob/master/SingleSource/Benchmarks/Linpack/linpack-pc.c" target="_blank">https://github.com/llvm/llvm-test-suite/blob/master/SingleSource/Benchmarks/Linpack/linpack-pc.c</a></div><div> <br></div></div></div></div></div></div></blockquote><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div></div><div>I'd very much appreciate if you could help me with this. Thank you.</div></div></div></div></div></div></blockquote><div><br></div><div>Your approach seems sensible to me. I usually have been debugging this kind of problem by piping the output of the two runs with `-print-after-all` to files and diff them to find out where the difference pops in.</div><div><br></div><div>-- </div><div>Mehdi</div><div><br></div></div></div>
_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a><br>
<a href="https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev" rel="noreferrer" target="_blank">https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a><br>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail-m_-4677688217354519596gmail-m_3753848777816909818gmail_signature"><div dir="ltr"><div><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div><font size="1"><span style="color:rgb(84,141,212);font-family:arial,helvetica,sans-serif">Regards</span><br></font></div><div><div dir="ltr"><font color="#000000" face="arial, helvetica, sans-serif" size="1"><font color="#548dd4">Bhatu</font></font></div></div></div></div></div></div></div></div></div></div>
</blockquote></div>