<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=iso-8859-1">
<meta name="Generator" content="Microsoft Word 14 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:Helvetica;
panose-1:2 11 6 4 2 2 2 2 2 4;}
@font-face
{font-family:Helvetica;
panose-1:2 11 6 4 2 2 2 2 2 4;}
@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: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;}
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:12.0pt;
font-family:"Times New Roman","serif";}
span.apple-converted-space
{mso-style-name:apple-converted-space;}
span.EmailStyle18
{mso-style-type:personal-reply;
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";}
.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;}
/* List Definitions */
@list l0
{mso-list-id:623266550;
mso-list-type:hybrid;
mso-list-template-ids:-1419372098 67698703 67698713 67698715 67698703 67698713 67698715 67698703 67698713 67698715;}
@list l0:level1
{mso-level-tab-stop:none;
mso-level-number-position:left;
text-indent:-18.0pt;}
@list l0:level2
{mso-level-number-format:alpha-lower;
mso-level-tab-stop:none;
mso-level-number-position:left;
text-indent:-18.0pt;}
@list l0:level3
{mso-level-number-format:roman-lower;
mso-level-tab-stop:none;
mso-level-number-position:right;
text-indent:-9.0pt;}
@list l0:level4
{mso-level-tab-stop:none;
mso-level-number-position:left;
text-indent:-18.0pt;}
@list l0:level5
{mso-level-number-format:alpha-lower;
mso-level-tab-stop:none;
mso-level-number-position:left;
text-indent:-18.0pt;}
@list l0:level6
{mso-level-number-format:roman-lower;
mso-level-tab-stop:none;
mso-level-number-position:right;
text-indent:-9.0pt;}
@list l0:level7
{mso-level-tab-stop:none;
mso-level-number-position:left;
text-indent:-18.0pt;}
@list l0:level8
{mso-level-number-format:alpha-lower;
mso-level-tab-stop:none;
mso-level-number-position:left;
text-indent:-18.0pt;}
@list l0:level9
{mso-level-number-format:roman-lower;
mso-level-tab-stop:none;
mso-level-number-position:right;
text-indent:-9.0pt;}
ol
{margin-bottom:0cm;}
ul
{margin-bottom:0cm;}
--></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">Hi Quentin,<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">I have rerun the test case on a recent commit, so the numbers have changed. There are also now a few more basic blocks very small basic blocks in the function,
and therefore there are some slight differences. I tried to go back to earlier commits, without success for some reason… This is however very similar, except that there becomes two COPYs back to sibling value after the loop. My apologies [vregs 76->111, 87->122].<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">1.
<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">The interval for %vreg111 first covers nearly the entire function. Then it gets split into two intervals, where one covers the inner loops, which makes sense.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">selectOrSplit %vreg111 [68r,400B:1)[400B,688r:6)[688r,752B:4)[752B,1264r:6)[1264r,1312r:3)[1312r,1472B:2)[1472B,1520r:5)[1520r,3488B:0) 0@1520r 1@68r 2@1312r
3@1264r 4@688r 5@1472B-phi 6@400B-phi w=3.181050e-02<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">queuing new interval: %vreg121 [1764r,2936r:0)[2960B,2980r:0) 0@1764r<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">queuing new interval: %vreg122 [68r,400B:0)[400B,688r:1)[688r,752B:2)[752B,1264r:1)[1264r,1312r:3)[1312r,1472B:4)[1472B,1520r:5)[1520r,1764r:6)[2936r,2960B:7)[2980r,2988B:8)[2988B,3024B:9)[3024B,3488B:10)
0@68r 1@400B-phi 2@688r 3@1264r 4@1312r 5@1472B-phi 6@1520r 7@2936r 8@2980r 9@2988B-phi 10@3024B-phi<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 lang="SV" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">2.
<o:p></o:p></span></p>
<p class="MsoNormal"><span lang="SV" style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Inline spilling %vreg121 [1764r,2948r:0)[2960B,2968r:0) 0@1764r<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">From original %vreg111<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Tracing value %vreg121:0@1764r<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> %vreg121:0@1764r: copy of %vreg122:6@1520r kill=1<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> %vreg122:6@1520r: copy of %vreg122:5@1472B kill=1<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> %vreg122:5@1472B: split phi value, checking 2 phi-defs, and 3 non-phi/orig defs // WRONG: the phi value was part of %vreg111 interval, see above.<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">The non-phi/defs to check are the sibling COPYs after the inner loops from %vreg121 to %vreg122 (@2948, @2968), and the identity COPY from valno 5 to 6 @1520
for %vreg122.<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">3.
<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Vreg122:5 @1472 is the beginning of a basic block. It was there all along, first in %vreg111, and then in %vreg122. %vreg111 gets split two blocks further down
@1764.<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">Outer loop<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">BB2:<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">BB15:<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">BB3: //joining BB2 and BB15<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> @1472 // PHI value that should have been detected as part of OrigLI.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> @1520 identity COPY 122 to 122<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">BB4:<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">BB5:<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> @1764 // This is where vreg111 is split and tracing of vreg121 begins. OrigVNI passed is @1520.<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"> // Exits with two edges into two one-block loops, with one exit block each, which are joined in BB12.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> // Succs BB6, BB8<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">BB6:
<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> {} // Single block nested loop<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">BB7: // pred BB6, succ BB12<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">@2948 sibling COPY 121 to 122<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">BB8: // pred BB5<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> {} // Single block nested loop<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">BB9: // pred BB8, succ BB12<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">@2968 sibling COPY 121 to 122<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">BB12: // joining BB7 and BB9<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">phi 9<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">BB13:<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> // Latch block for outer loop, has edge from BB#4<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> phi 10<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>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Is there anything else I can provide that might you?<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">/Jonas<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 0cm 0cm 0cm">
<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""> Quentin Colombet [mailto:qcolombet@apple.com]
<br>
<b>Sent:</b> den 2 december 2014 01:15<br>
<b>To:</b> Jonas Paulsson<br>
<b>Cc:</b> llvmdev@cs.uiuc.edu; stoklund@2pi.dk; Patrik Hägglund H<br>
<b>Subject:</b> Re: [LLVMdev] InlineSpiller.cpp bug?<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<p class="MsoNormal">Hi Jonas,<o:p></o:p></p>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">Thanks for your patience.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">I have looked into the problem you reported and although the fix you proposed seem correct, I am not sure yet this is the way to go. I would need the debug output of the regalloc to help you further, but here are my initial findings.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">traceSiblingValue </span>looked into the problematic phi because the previous iteration said that <span style="font-size:11.0pt;font-family:"Courier New"">%vreg87:6
</span>and <span style="font-size:11.0pt;font-family:"Courier New"">%vreg87:5 </span>are sibling.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">Since they are sibling it means they share the same original vreg.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">The problem is that if vreg87:5, i.e., the original phi, is not the original value then it must have been inserted by splitting.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">Based on what you said this is not the case, so we end up in this weird situation.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">If you cannot share the debug output, you could check the following for me:<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">1. how vreg76 gets split?<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">2. what is the second non-phi/defs for vreg87:5?<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">3. if vreg87:5 is not inserted by splitting, how does it gets here?<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal">Thanks,<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">-Quentin<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>
<p class="MsoNormal">On Nov 21, 2014, at 9:22 AM, Jonas Paulsson <<a href="mailto:jonas.paulsson@ericsson.com">jonas.paulsson@ericsson.com</a>> wrote:<o:p></o:p></p>
</div>
<p class="MsoNormal"><br>
<br>
<o:p></o:p></p>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Hi Quentin,</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">I have tried to find a test case for an official target, but failed. It seems to be a rare case.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">To do it, I added the ‘else’ clause in the following:</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">…</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">if (VNI->def == OrigVNI->def) {</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> DEBUG(dbgs() << "orig phi value\n");</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> SVI->second.DefByOrigPHI = true;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> SVI->second.AllDefsAreReloads = false;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> propagateSiblingValue(SVI);</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> continue;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">}</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> //</span><span class="apple-converted-space"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> </span></span><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">check
if the valno is actually an orig PHI, but is not OrigVNI</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">else {</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> LiveInterval &OrigLI = LIS.getInterval(Original);</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> VNInfo *OrigVNI_curr = OrigLI.getVNInfoAt(VNI->def);</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> if (OrigVNI_curr->def == VNI->def)</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> assert(0 && "OrigLI contained VNI which was a PHI, but not OrigVNI!");</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">}</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">, but the assert never triggered anywhere else than in my original case.</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">When I did the following, at least my test case worked correctly:</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">LiveInterval &OrigLI = LIS.getInterval(Original);</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">VNInfo *OrigVNI_curr = OrigLI.getVNInfoAt(VNI->def);</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">if (OrigVNI_curr->def == VNI->def) {</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> DEBUG(dbgs() << "orig phi value\n");</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> SVI->second.DefByOrigPHI = true;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> SVI->second.AllDefsAreReloads = false;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> propagateSiblingValue(SVI);</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D"> continue;</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New";color:#1F497D">}</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">The question is: is this sound and safe?</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">It seems logical to me, but I would really appreciate some expert advice,</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Jonas</span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"> </span><o:p></o:p></p>
</div>
<div>
<div style="border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<div>
<p class="MsoNormal"><b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">From:</span></b><span class="apple-converted-space"><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif""> </span></span><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">Quentin
Colombet [<a href="mailto:qcolombet@apple.com">mailto:qcolombet@apple.com</a>]<span class="apple-converted-space"> </span><br>
<b>Sent:</b><span class="apple-converted-space"> </span>den 18 november 2014 20:19<br>
<b>To:</b><span class="apple-converted-space"> </span>Jonas Paulsson<br>
<b>Cc:</b><span class="apple-converted-space"> </span><a href="mailto:llvmdev@cs.uiuc.edu">llvmdev@cs.uiuc.edu</a>;
<a href="mailto:stoklund@2pi.dk">stoklund@2pi.dk</a><br>
<b>Subject:</b><span class="apple-converted-space"> </span>Re: [LLVMdev] InlineSpiller.cpp bug?</span><o:p></o:p></p>
</div>
</div>
</div>
<div>
<p class="MsoNormal"> <o:p></o:p></p>
</div>
<div>
<p class="MsoNormal">Hi Jonas,<o:p></o:p></p>
</div>
<div>
<div>
<p class="MsoNormal"> <o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal">Could you share your test case and/or file a PR?<o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"> <o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal">I’d like to see what is going on dynamically to help you.<o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"> <o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal">Thanks,<o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal">-Quentin<o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"> <o:p></o:p></p>
</div>
<div>
<div>
<div>
<p class="MsoNormal">On Nov 18, 2014, at 3:26 AM, Jonas Paulsson <<a href="mailto:jonas.paulsson@ericsson.com"><span style="color:purple">jonas.paulsson@ericsson.com</span></a>> wrote:<o:p></o:p></p>
</div>
</div>
<div>
<p class="MsoNormal"><br>
<br>
<br>
<o:p></o:p></p>
</div>
<div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Hi,</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">I have encountered a test case where InlineSpiller generates bad code. A register is reloaded but never spilled, and I suspect a bug in InlineSpiller.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">A register is live over a loop that also have two inner loops. It is not used or defined over the inner loops. It is split into two sibling registers, where one covers just
the inner loops interval, which is then spilled.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">In spill(), analyzeSiblingValues() is called, which calls traceSiblingValue(). It traces in several iterations strangely back to the same register (inside a loop), finds
it marked to be spilled, and the spill is cancelled:</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span lang="SV" style="font-size:11.0pt;font-family:"Courier New"">Inline spilling %vreg86 [1396r,2276r:0) 0@1396r</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New"">From original %vreg76</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New"">Tracing value %vreg86:0@1396r</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New""> %vreg86:0@1396r: copy of %vreg87:6@1168r kill=1</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New""> %vreg87:6@1168r: copy of %vreg87:5@1120B kill=1</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New""> %vreg87:5@1120B: split phi value, checking 1 phi-defs, and 2 non-phi/orig defs</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New""> %vreg87:7@2276r: copy of %vreg86:0@1396r kill=1</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New""> traced to: spill %vreg86:0@1396r all-reloads kill deps[ 7@2276r ]</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New"">Merged spilled regs: SS#1 [1396r,2276r:0) 0@x</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">I am guessing that traceSiblingValue() should have stopped at a PHI ValNo by recognizing it as ‘original’, meaning it was not inserted by splitting. It does not although
it is clear that this PHI ValNo is part of OrigLI.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">This is how it looked, roughly:</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Original LiveInterval:</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">5 0 --- inner loops --- // last valno in interval</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">PHI COPY of valno 5</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">After splitting:</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">5 6 7 8</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">PHI COPY of valno 5 COPY of valno 0 PHI</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> 0</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> COPY of valno 6</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> /\</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> OrigVNI</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Tracing sibling values, valno 6 is the original valno:</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Valno 0 is a copy from 6.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Valno 6 is a copy from 5.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Valno 5 is a phi. Is it OrigVNI? NO! ‘Therefore it is not an original phi.’ WRONG! The search continues past it, assuming that it is a newly inserted PHI, done during splitting.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">My conlusion is that either the line</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Courier New"">if (VNI->def == OrigVNI->def)</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">is wrong, because it doesn’t really check that VNI was not a phi in OrigLI, because it assumes that if VNI is a phi which was part of the original LI, then OrigVNI must be
that PHI. This is not the case here.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">The algorithm has iterated past OrigVNI and VNI is at this point a phi that was part of the original LI, which is the copy source of OrigVNI.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Or, it is assumed that splitting is always done at PHI ValNos somehow, and not as in this case, by a COPY of a PHI ValNo.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">I would very much appreciate assistance in resolving this problem. As explained above, it is not clear to me why this error occurs, or what is the appropriate fix. I can
provide more details if needed for some reason.</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Best regards,</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif"">Jonas Paulsson</span><o:p></o:p></p>
</div>
</div>
<div>
<div>
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif""> </span><o:p></o:p></p>
</div>
</div>
<div>
<p class="MsoNormal"><span style="font-size:9.0pt;font-family:"Helvetica","sans-serif"">_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:LLVMdev@cs.uiuc.edu"><span style="color:purple">LLVMdev@cs.uiuc.edu</span></a><span class="apple-converted-space"> </span> <a href="http://llvm.cs.uiuc.edu/"><span style="color:purple">http://llvm.cs.uiuc.edu</span></a><br>
<a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev"><span style="color:purple">http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev</span></a></span><o:p></o:p></p>
</div>
</div>
</div>
</div>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
</div>
</div>
</body>
</html>