<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 12 (filtered medium)"><base href="x-msg://10814/"><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:Consolas;
panose-1:2 11 6 9 2 2 4 3 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;}
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";}
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: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'><o:p> </o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Andy and everyone interested, <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'> This patch addresses the MI DAG construction indeterminism discussed in the thread below. It also needs a new method (find()) for MapVector<>.<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>It tests clean on Hexagon and x86.<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'>Please review.<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'>Sergei<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><p class=MsoNormal><span style='font-size:10.5pt;font-family:Consolas;color:#1F497D'>---<o:p></o:p></span></p><p class=MsoNormal><span style='font-size:10.5pt;font-family:"Calibri","sans-serif";color:#1F497D'>Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation</span><span style='font-size:10.5pt;font-family:Consolas;color:#1F497D'><o:p></o:p></span></p></div><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"'> Andrew Trick [mailto:atrick@apple.com] <br><b>Sent:</b> Wednesday, October 17, 2012 5:57 PM<br><b>To:</b> Sergei Larin<br><b>Cc:</b> 'LLVM Developers Mailing List'<br><b>Subject:</b> Re: [LLVMdev] MI DAG constructor indeterminism<o:p></o:p></span></p></div></div><p class=MsoNormal><o:p> </o:p></p><p class=MsoNormal><o:p> </o:p></p><div><div><p class=MsoNormal>On Oct 17, 2012, at 3:10 PM, Sergei Larin <<a href="mailto:slarin@codeaurora.org">slarin@codeaurora.org</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'>Andy,</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'> So if it is not a feature… then couple questions:</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'> First, I also do not see an easy way to restructure work sets in this case – so let’s assume std::map is needed here. Then the way I understand it, there are five objects that cause the indeterminism:</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'> std::map<const Value *, SUnit *> AliasMemDefs, NonAliasMemDefs;</span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> std::map<const Value *, std::vector<SUnit *> > AliasMemUses, NonAliasMemUses;</span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> std::set<SUnit*> RejectMemNodes;</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'>…since all of them at different point of time are traversed begin to end, and as such are affected by pointer value.</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 case of “std::set<SUnit*> RejectMemNodes; “ is easy. Something like this will work:</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'>struct SortSUComp {</span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> bool operator()(const SUnit *left, const SUnit *right) const {</span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> return left->NodeNum < right->NodeNum;</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:"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'> typedef std::set<SUnit*, SortSUComp> SortedSUSet;</span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> SortedSUSet RejectMemNodes;</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 in case of Value * as a key – is there any unique ID that is deterministic and could be used here? As I understand getValueID () is not unique, or could it be sufficient in this 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'>Something as bulky as temporary sorting queue/vector is also possible, but I do not like it very much.</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'>Any other ideas?</span><o:p></o:p></p></div></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>I don't have any great ideas, but anything is an improvement over std::map/std::set.<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>AFAIK we don't need to remove individual elements from the map. Is that right? So SetVector and MapVector should work fine. They don't sort by key but maintain a side vector to track insertion order.<o:p></o:p></p></div><div><p class=MsoNormal><o:p> </o:p></p></div><div><p class=MsoNormal>-Andy<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'> </span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Thanks.</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'>Sergei</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><p class=MsoNormal><span style='font-size:10.5pt;font-family:Consolas;color:#1F497D'>---</span><o:p></o:p></p></div><div><p class=MsoNormal><span style='font-size:10.5pt;font-family:"Calibri","sans-serif";color:#1F497D'>Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by The Linux Foundation</span><o:p></o:p></p></div></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 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'><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"'>Andrew Trick [mailto:atrick@<a href="http://apple.com">apple.com</a>]<span class=apple-converted-space> </span><br><b>Sent:</b><span class=apple-converted-space> </span>Tuesday, October 16, 2012 10:44 PM<br><b>To:</b><span class=apple-converted-space> </span>Sergei Larin<br><b>Cc:</b><span class=apple-converted-space> </span>'LLVM Developers Mailing List'<br><b>Subject:</b><span class=apple-converted-space> </span>Re: [LLVMdev] MI DAG constructor indeterminism</span><o:p></o:p></p></div></div></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 Oct 16, 2012, at 1:43 PM, Sergei Larin <<a href="mailto:slarin@codeaurora.org"><span style='color:purple'>slarin@codeaurora.org</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";color:#1F497D'> </span><o:p></o:p></p></div></div><div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Andy,</span><o:p></o:p></p></div></div><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><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> This is less of a question but rather a status quo verification…</span><o:p></o:p></p></div></div><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><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> We currently have certain indeterminism in MI scheduler DAG construction – it is introduces by the use of std::map/std::set during edge traversal.</span><o:p></o:p></p></div></div><div><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'>Result – a random variation in SUnit edge order (which will remain fixed thereafter). Logically, it is the same DAG, but topologically it is a slightly different one, and if some algorithm is dependent on the order of edge traversal, we can have performance and debugging indeterminism. The way I have discovered it – VLIW scheduler can produce identical cost function for a pair of SUs, making visitation order the tie breaker, which is not deterministic per above discussion. For me it is trivial to fix, but I wonder if this might become a source of well hidden issues in the future.</span><o:p></o:p></p></div></div><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><div><p class=MsoNormal><span style='font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D'> I am at this time not proposing anything – a fix is definitely possible, but I wonder what people think about it before I even consider this a bug.</span><o:p></o:p></p></div></div></div><div><p class=MsoNormal> <o:p></o:p></p></div></div><div><div><p class=MsoNormal>This looks like a bug. The edge order could even affect some heuristics and influence codegen.<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 don't have a better idea than using SetVector/MapVector for Value* keys. For SUnits we could just key on NodeNum. Go ahead and file a bug and/or submit a patch.<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>-Andy<o:p></o:p></p></div></div></div></div></div><p class=MsoNormal><o:p> </o:p></p></div></div></body></html>