<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Jul 15, 2015 at 5:34 PM, Chandler Carruth <span dir="ltr"><<a href="mailto:chandlerc@google.com" target="_blank">chandlerc@google.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div class="gmail_quote"><span class=""><div dir="ltr">On Wed, Jul 15, 2015 at 12:55 PM Hyojin Sung <<a href="mailto:hsung@us.ibm.com" target="_blank">hsung@us.ibm.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div>
<p><font size="2" face="Helvetica">Hi all,</font><br>
<br>
<font size="2" face="Helvetica">I would like to propose an improvement of the “almost dead” block elimination in Transforms/Local.cpp so that it will preserve the canonical loop form for loops with a volatile iteration variable.</font><br>
<br>
<font size="2" face="Helvetica">*** Problem statement</font><br>
<font size="2" face="Helvetica">Nested loops in LCALS Subset B (</font><a href="https://urldefense.proofpoint.com/v2/url?u=https-3A__codesign.llnl.gov_LCALS.php&d=AwMGaQ&c=8hUWFZcy2Z-Za5rBPlktOQ&r=Mfk2qtn1LTDThVkh6-oGglNfMADXfJdty4_bhmuhMHA&m=aWKfvN4c8lvUSvVn8J0Z2ajTctlBJf0198Au28epBr0&s=4d9dt5ODcDWHHatSrwu5ZYT9ebgVzNEtpOlIR87izCM&e=" target="_blank"><font size="2" face="Helvetica"><u>https://codesign.llnl.gov/LCALS.php</u></font></a><font size="2" face="Helvetica">) are not vectorized with LLVM -O3 because the LLVM loop vectorizer fails the test whether the loop latch and exiting block of a loop is the same. The loops are vectorizable, and get vectorized with LLVM -O2</font></p></div></blockquote></span><div>I would be interested to know why -O2 succeeds here.</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><p><font size="2" face="Helvetica"> and also with other commercial compilers (icc, xlc). </font><br>
<br>
<font size="2" face="Helvetica">*** Details</font><br>
<font size="2" face="Helvetica">These loops ended up with different loop latch and exiting block after a series of optimizations including loop unswitching, jump threading, simplify-the-CFG, and loop simplify. The fundamental problem here is that the above optimizations cannot recognize a loop with a volatile iteration variable and do not preserve its canonical loop structure.</font></p></div></blockquote></span><div>Ok, meta-level question first:</div><div><br></div><div>Why do we care about performance of loops with a volatile iteration variable? That seems both counter-intuitive and unlikely to be a useful goal. We simply don't optimize volatile operations well in *any* part of the optimizer, and I'm not sure why we need to start trying to fix that. This seems like an irreparably broken benchmark, but perhaps there is a motivation I don't yet see.</div></div></div></blockquote><div><br></div><div>A quick look at the tarball on the linked site suggests that the volatile iteration variable is done on purpose so that the outer "run thing N times and take the average" loop can't be optimized.</div><div><br></div><div>-- Sean Silva</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div class="gmail_quote"><div><br></div><div><br></div><div>Assuming that sufficient motivation arises to try to fix this, see my comments below:</div><span class=""><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><p><br>
<br>
<font size="2" face="Helvetica">(1) Loop unswitching generates several empty placeholder BBs only with PHI nodes after separating out a shorter path with no inner loop execution from a standard path. </font><br>
<br>
<font size="2" face="Helvetica">(2) Jump threading and simplify-the-CFG passes independently calls TryToSimplifyUnconditionalBranchFromEmptyBlock() in Transforms/Utils/Local.cpp to get rid of almost empty BBs. </font><br>
<br>
<font size="2" face="Helvetica">(3) TryToSimplifyUnconditionalBranchFromEmtpyBlock() eliminates the placeholder BBs after loop unswitching and merges them into subsequent blocks including the header of the inner loop. Before eliminating the blocks, the function checks if the block is a loop header by looking at its PHI nodes so that it can be saved, but the test fails with the loops with a volatile iteration variable.</font></p></div></blockquote></span><div>Why does this fail for a volatile iteration variable but not for a non-volatile one? I think understanding that will be key to understanding how it should be fixed.</div></div></div>
<br>_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:LLVMdev@cs.uiuc.edu">LLVMdev@cs.uiuc.edu</a> <a href="http://llvm.cs.uiuc.edu" rel="noreferrer" target="_blank">http://llvm.cs.uiuc.edu</a><br>
<a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev" rel="noreferrer" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev</a><br>
<br></blockquote></div><br></div></div>