xref: /freebsd/contrib/llvm-project/llvm/lib/Analysis/BlockFrequencyInfoImpl.cpp (revision ec0ea6efa1ad229d75c394c1a9b9cac33af2b1d3)
1 //===- BlockFrequencyImplInfo.cpp - Block Frequency Info Implementation ---===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // Loops should be simplified before this analysis.
10 //
11 //===----------------------------------------------------------------------===//
12 
13 #include "llvm/Analysis/BlockFrequencyInfoImpl.h"
14 #include "llvm/ADT/APInt.h"
15 #include "llvm/ADT/DenseMap.h"
16 #include "llvm/ADT/GraphTraits.h"
17 #include "llvm/ADT/None.h"
18 #include "llvm/ADT/SCCIterator.h"
19 #include "llvm/Config/llvm-config.h"
20 #include "llvm/IR/Function.h"
21 #include "llvm/Support/BlockFrequency.h"
22 #include "llvm/Support/BranchProbability.h"
23 #include "llvm/Support/Compiler.h"
24 #include "llvm/Support/Debug.h"
25 #include "llvm/Support/ScaledNumber.h"
26 #include "llvm/Support/MathExtras.h"
27 #include "llvm/Support/raw_ostream.h"
28 #include <algorithm>
29 #include <cassert>
30 #include <cstddef>
31 #include <cstdint>
32 #include <iterator>
33 #include <list>
34 #include <numeric>
35 #include <utility>
36 #include <vector>
37 
38 using namespace llvm;
39 using namespace llvm::bfi_detail;
40 
41 #define DEBUG_TYPE "block-freq"
42 
43 namespace llvm {
44 cl::opt<bool> CheckBFIUnknownBlockQueries(
45     "check-bfi-unknown-block-queries",
46     cl::init(false), cl::Hidden,
47     cl::desc("Check if block frequency is queried for an unknown block "
48              "for debugging missed BFI updates"));
49 
50 cl::opt<bool> UseIterativeBFIInference(
51     "use-iterative-bfi-inference", cl::init(false), cl::Hidden, cl::ZeroOrMore,
52     cl::desc("Apply an iterative post-processing to infer correct BFI counts"));
53 
54 cl::opt<unsigned> IterativeBFIMaxIterationsPerBlock(
55     "iterative-bfi-max-iterations-per-block", cl::init(1000), cl::Hidden,
56     cl::desc("Iterative inference: maximum number of update iterations "
57              "per block"));
58 
59 cl::opt<double> IterativeBFIPrecision(
60     "iterative-bfi-precision", cl::init(1e-12), cl::Hidden,
61     cl::desc("Iterative inference: delta convergence precision; smaller values "
62              "typically lead to better results at the cost of worsen runtime"));
63 }
64 
65 ScaledNumber<uint64_t> BlockMass::toScaled() const {
66   if (isFull())
67     return ScaledNumber<uint64_t>(1, 0);
68   return ScaledNumber<uint64_t>(getMass() + 1, -64);
69 }
70 
71 #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
72 LLVM_DUMP_METHOD void BlockMass::dump() const { print(dbgs()); }
73 #endif
74 
75 static char getHexDigit(int N) {
76   assert(N < 16);
77   if (N < 10)
78     return '0' + N;
79   return 'a' + N - 10;
80 }
81 
82 raw_ostream &BlockMass::print(raw_ostream &OS) const {
83   for (int Digits = 0; Digits < 16; ++Digits)
84     OS << getHexDigit(Mass >> (60 - Digits * 4) & 0xf);
85   return OS;
86 }
87 
88 namespace {
89 
90 using BlockNode = BlockFrequencyInfoImplBase::BlockNode;
91 using Distribution = BlockFrequencyInfoImplBase::Distribution;
92 using WeightList = BlockFrequencyInfoImplBase::Distribution::WeightList;
93 using Scaled64 = BlockFrequencyInfoImplBase::Scaled64;
94 using LoopData = BlockFrequencyInfoImplBase::LoopData;
95 using Weight = BlockFrequencyInfoImplBase::Weight;
96 using FrequencyData = BlockFrequencyInfoImplBase::FrequencyData;
97 
98 /// Dithering mass distributer.
99 ///
100 /// This class splits up a single mass into portions by weight, dithering to
101 /// spread out error.  No mass is lost.  The dithering precision depends on the
102 /// precision of the product of \a BlockMass and \a BranchProbability.
103 ///
104 /// The distribution algorithm follows.
105 ///
106 ///  1. Initialize by saving the sum of the weights in \a RemWeight and the
107 ///     mass to distribute in \a RemMass.
108 ///
109 ///  2. For each portion:
110 ///
111 ///      1. Construct a branch probability, P, as the portion's weight divided
112 ///         by the current value of \a RemWeight.
113 ///      2. Calculate the portion's mass as \a RemMass times P.
114 ///      3. Update \a RemWeight and \a RemMass at each portion by subtracting
115 ///         the current portion's weight and mass.
116 struct DitheringDistributer {
117   uint32_t RemWeight;
118   BlockMass RemMass;
119 
120   DitheringDistributer(Distribution &Dist, const BlockMass &Mass);
121 
122   BlockMass takeMass(uint32_t Weight);
123 };
124 
125 } // end anonymous namespace
126 
127 DitheringDistributer::DitheringDistributer(Distribution &Dist,
128                                            const BlockMass &Mass) {
129   Dist.normalize();
130   RemWeight = Dist.Total;
131   RemMass = Mass;
132 }
133 
134 BlockMass DitheringDistributer::takeMass(uint32_t Weight) {
135   assert(Weight && "invalid weight");
136   assert(Weight <= RemWeight);
137   BlockMass Mass = RemMass * BranchProbability(Weight, RemWeight);
138 
139   // Decrement totals (dither).
140   RemWeight -= Weight;
141   RemMass -= Mass;
142   return Mass;
143 }
144 
145 void Distribution::add(const BlockNode &Node, uint64_t Amount,
146                        Weight::DistType Type) {
147   assert(Amount && "invalid weight of 0");
148   uint64_t NewTotal = Total + Amount;
149 
150   // Check for overflow.  It should be impossible to overflow twice.
151   bool IsOverflow = NewTotal < Total;
152   assert(!(DidOverflow && IsOverflow) && "unexpected repeated overflow");
153   DidOverflow |= IsOverflow;
154 
155   // Update the total.
156   Total = NewTotal;
157 
158   // Save the weight.
159   Weights.push_back(Weight(Type, Node, Amount));
160 }
161 
162 static void combineWeight(Weight &W, const Weight &OtherW) {
163   assert(OtherW.TargetNode.isValid());
164   if (!W.Amount) {
165     W = OtherW;
166     return;
167   }
168   assert(W.Type == OtherW.Type);
169   assert(W.TargetNode == OtherW.TargetNode);
170   assert(OtherW.Amount && "Expected non-zero weight");
171   if (W.Amount > W.Amount + OtherW.Amount)
172     // Saturate on overflow.
173     W.Amount = UINT64_MAX;
174   else
175     W.Amount += OtherW.Amount;
176 }
177 
178 static void combineWeightsBySorting(WeightList &Weights) {
179   // Sort so edges to the same node are adjacent.
180   llvm::sort(Weights, [](const Weight &L, const Weight &R) {
181     return L.TargetNode < R.TargetNode;
182   });
183 
184   // Combine adjacent edges.
185   WeightList::iterator O = Weights.begin();
186   for (WeightList::const_iterator I = O, L = O, E = Weights.end(); I != E;
187        ++O, (I = L)) {
188     *O = *I;
189 
190     // Find the adjacent weights to the same node.
191     for (++L; L != E && I->TargetNode == L->TargetNode; ++L)
192       combineWeight(*O, *L);
193   }
194 
195   // Erase extra entries.
196   Weights.erase(O, Weights.end());
197 }
198 
199 static void combineWeightsByHashing(WeightList &Weights) {
200   // Collect weights into a DenseMap.
201   using HashTable = DenseMap<BlockNode::IndexType, Weight>;
202 
203   HashTable Combined(NextPowerOf2(2 * Weights.size()));
204   for (const Weight &W : Weights)
205     combineWeight(Combined[W.TargetNode.Index], W);
206 
207   // Check whether anything changed.
208   if (Weights.size() == Combined.size())
209     return;
210 
211   // Fill in the new weights.
212   Weights.clear();
213   Weights.reserve(Combined.size());
214   for (const auto &I : Combined)
215     Weights.push_back(I.second);
216 }
217 
218 static void combineWeights(WeightList &Weights) {
219   // Use a hash table for many successors to keep this linear.
220   if (Weights.size() > 128) {
221     combineWeightsByHashing(Weights);
222     return;
223   }
224 
225   combineWeightsBySorting(Weights);
226 }
227 
228 static uint64_t shiftRightAndRound(uint64_t N, int Shift) {
229   assert(Shift >= 0);
230   assert(Shift < 64);
231   if (!Shift)
232     return N;
233   return (N >> Shift) + (UINT64_C(1) & N >> (Shift - 1));
234 }
235 
236 void Distribution::normalize() {
237   // Early exit for termination nodes.
238   if (Weights.empty())
239     return;
240 
241   // Only bother if there are multiple successors.
242   if (Weights.size() > 1)
243     combineWeights(Weights);
244 
245   // Early exit when combined into a single successor.
246   if (Weights.size() == 1) {
247     Total = 1;
248     Weights.front().Amount = 1;
249     return;
250   }
251 
252   // Determine how much to shift right so that the total fits into 32-bits.
253   //
254   // If we shift at all, shift by 1 extra.  Otherwise, the lower limit of 1
255   // for each weight can cause a 32-bit overflow.
256   int Shift = 0;
257   if (DidOverflow)
258     Shift = 33;
259   else if (Total > UINT32_MAX)
260     Shift = 33 - countLeadingZeros(Total);
261 
262   // Early exit if nothing needs to be scaled.
263   if (!Shift) {
264     // If we didn't overflow then combineWeights() shouldn't have changed the
265     // sum of the weights, but let's double-check.
266     assert(Total == std::accumulate(Weights.begin(), Weights.end(), UINT64_C(0),
267                                     [](uint64_t Sum, const Weight &W) {
268                       return Sum + W.Amount;
269                     }) &&
270            "Expected total to be correct");
271     return;
272   }
273 
274   // Recompute the total through accumulation (rather than shifting it) so that
275   // it's accurate after shifting and any changes combineWeights() made above.
276   Total = 0;
277 
278   // Sum the weights to each node and shift right if necessary.
279   for (Weight &W : Weights) {
280     // Scale down below UINT32_MAX.  Since Shift is larger than necessary, we
281     // can round here without concern about overflow.
282     assert(W.TargetNode.isValid());
283     W.Amount = std::max(UINT64_C(1), shiftRightAndRound(W.Amount, Shift));
284     assert(W.Amount <= UINT32_MAX);
285 
286     // Update the total.
287     Total += W.Amount;
288   }
289   assert(Total <= UINT32_MAX);
290 }
291 
292 void BlockFrequencyInfoImplBase::clear() {
293   // Swap with a default-constructed std::vector, since std::vector<>::clear()
294   // does not actually clear heap storage.
295   std::vector<FrequencyData>().swap(Freqs);
296   IsIrrLoopHeader.clear();
297   std::vector<WorkingData>().swap(Working);
298   Loops.clear();
299 }
300 
301 /// Clear all memory not needed downstream.
302 ///
303 /// Releases all memory not used downstream.  In particular, saves Freqs.
304 static void cleanup(BlockFrequencyInfoImplBase &BFI) {
305   std::vector<FrequencyData> SavedFreqs(std::move(BFI.Freqs));
306   SparseBitVector<> SavedIsIrrLoopHeader(std::move(BFI.IsIrrLoopHeader));
307   BFI.clear();
308   BFI.Freqs = std::move(SavedFreqs);
309   BFI.IsIrrLoopHeader = std::move(SavedIsIrrLoopHeader);
310 }
311 
312 bool BlockFrequencyInfoImplBase::addToDist(Distribution &Dist,
313                                            const LoopData *OuterLoop,
314                                            const BlockNode &Pred,
315                                            const BlockNode &Succ,
316                                            uint64_t Weight) {
317   if (!Weight)
318     Weight = 1;
319 
320   auto isLoopHeader = [&OuterLoop](const BlockNode &Node) {
321     return OuterLoop && OuterLoop->isHeader(Node);
322   };
323 
324   BlockNode Resolved = Working[Succ.Index].getResolvedNode();
325 
326 #ifndef NDEBUG
327   auto debugSuccessor = [&](const char *Type) {
328     dbgs() << "  =>"
329            << " [" << Type << "] weight = " << Weight;
330     if (!isLoopHeader(Resolved))
331       dbgs() << ", succ = " << getBlockName(Succ);
332     if (Resolved != Succ)
333       dbgs() << ", resolved = " << getBlockName(Resolved);
334     dbgs() << "\n";
335   };
336   (void)debugSuccessor;
337 #endif
338 
339   if (isLoopHeader(Resolved)) {
340     LLVM_DEBUG(debugSuccessor("backedge"));
341     Dist.addBackedge(Resolved, Weight);
342     return true;
343   }
344 
345   if (Working[Resolved.Index].getContainingLoop() != OuterLoop) {
346     LLVM_DEBUG(debugSuccessor("  exit  "));
347     Dist.addExit(Resolved, Weight);
348     return true;
349   }
350 
351   if (Resolved < Pred) {
352     if (!isLoopHeader(Pred)) {
353       // If OuterLoop is an irreducible loop, we can't actually handle this.
354       assert((!OuterLoop || !OuterLoop->isIrreducible()) &&
355              "unhandled irreducible control flow");
356 
357       // Irreducible backedge.  Abort.
358       LLVM_DEBUG(debugSuccessor("abort!!!"));
359       return false;
360     }
361 
362     // If "Pred" is a loop header, then this isn't really a backedge; rather,
363     // OuterLoop must be irreducible.  These false backedges can come only from
364     // secondary loop headers.
365     assert(OuterLoop && OuterLoop->isIrreducible() && !isLoopHeader(Resolved) &&
366            "unhandled irreducible control flow");
367   }
368 
369   LLVM_DEBUG(debugSuccessor(" local  "));
370   Dist.addLocal(Resolved, Weight);
371   return true;
372 }
373 
374 bool BlockFrequencyInfoImplBase::addLoopSuccessorsToDist(
375     const LoopData *OuterLoop, LoopData &Loop, Distribution &Dist) {
376   // Copy the exit map into Dist.
377   for (const auto &I : Loop.Exits)
378     if (!addToDist(Dist, OuterLoop, Loop.getHeader(), I.first,
379                    I.second.getMass()))
380       // Irreducible backedge.
381       return false;
382 
383   return true;
384 }
385 
386 /// Compute the loop scale for a loop.
387 void BlockFrequencyInfoImplBase::computeLoopScale(LoopData &Loop) {
388   // Compute loop scale.
389   LLVM_DEBUG(dbgs() << "compute-loop-scale: " << getLoopName(Loop) << "\n");
390 
391   // Infinite loops need special handling. If we give the back edge an infinite
392   // mass, they may saturate all the other scales in the function down to 1,
393   // making all the other region temperatures look exactly the same. Choose an
394   // arbitrary scale to avoid these issues.
395   //
396   // FIXME: An alternate way would be to select a symbolic scale which is later
397   // replaced to be the maximum of all computed scales plus 1. This would
398   // appropriately describe the loop as having a large scale, without skewing
399   // the final frequency computation.
400   const Scaled64 InfiniteLoopScale(1, 12);
401 
402   // LoopScale == 1 / ExitMass
403   // ExitMass == HeadMass - BackedgeMass
404   BlockMass TotalBackedgeMass;
405   for (auto &Mass : Loop.BackedgeMass)
406     TotalBackedgeMass += Mass;
407   BlockMass ExitMass = BlockMass::getFull() - TotalBackedgeMass;
408 
409   // Block scale stores the inverse of the scale. If this is an infinite loop,
410   // its exit mass will be zero. In this case, use an arbitrary scale for the
411   // loop scale.
412   Loop.Scale =
413       ExitMass.isEmpty() ? InfiniteLoopScale : ExitMass.toScaled().inverse();
414 
415   LLVM_DEBUG(dbgs() << " - exit-mass = " << ExitMass << " ("
416                     << BlockMass::getFull() << " - " << TotalBackedgeMass
417                     << ")\n"
418                     << " - scale = " << Loop.Scale << "\n");
419 }
420 
421 /// Package up a loop.
422 void BlockFrequencyInfoImplBase::packageLoop(LoopData &Loop) {
423   LLVM_DEBUG(dbgs() << "packaging-loop: " << getLoopName(Loop) << "\n");
424 
425   // Clear the subloop exits to prevent quadratic memory usage.
426   for (const BlockNode &M : Loop.Nodes) {
427     if (auto *Loop = Working[M.Index].getPackagedLoop())
428       Loop->Exits.clear();
429     LLVM_DEBUG(dbgs() << " - node: " << getBlockName(M.Index) << "\n");
430   }
431   Loop.IsPackaged = true;
432 }
433 
434 #ifndef NDEBUG
435 static void debugAssign(const BlockFrequencyInfoImplBase &BFI,
436                         const DitheringDistributer &D, const BlockNode &T,
437                         const BlockMass &M, const char *Desc) {
438   dbgs() << "  => assign " << M << " (" << D.RemMass << ")";
439   if (Desc)
440     dbgs() << " [" << Desc << "]";
441   if (T.isValid())
442     dbgs() << " to " << BFI.getBlockName(T);
443   dbgs() << "\n";
444 }
445 #endif
446 
447 void BlockFrequencyInfoImplBase::distributeMass(const BlockNode &Source,
448                                                 LoopData *OuterLoop,
449                                                 Distribution &Dist) {
450   BlockMass Mass = Working[Source.Index].getMass();
451   LLVM_DEBUG(dbgs() << "  => mass:  " << Mass << "\n");
452 
453   // Distribute mass to successors as laid out in Dist.
454   DitheringDistributer D(Dist, Mass);
455 
456   for (const Weight &W : Dist.Weights) {
457     // Check for a local edge (non-backedge and non-exit).
458     BlockMass Taken = D.takeMass(W.Amount);
459     if (W.Type == Weight::Local) {
460       Working[W.TargetNode.Index].getMass() += Taken;
461       LLVM_DEBUG(debugAssign(*this, D, W.TargetNode, Taken, nullptr));
462       continue;
463     }
464 
465     // Backedges and exits only make sense if we're processing a loop.
466     assert(OuterLoop && "backedge or exit outside of loop");
467 
468     // Check for a backedge.
469     if (W.Type == Weight::Backedge) {
470       OuterLoop->BackedgeMass[OuterLoop->getHeaderIndex(W.TargetNode)] += Taken;
471       LLVM_DEBUG(debugAssign(*this, D, W.TargetNode, Taken, "back"));
472       continue;
473     }
474 
475     // This must be an exit.
476     assert(W.Type == Weight::Exit);
477     OuterLoop->Exits.push_back(std::make_pair(W.TargetNode, Taken));
478     LLVM_DEBUG(debugAssign(*this, D, W.TargetNode, Taken, "exit"));
479   }
480 }
481 
482 static void convertFloatingToInteger(BlockFrequencyInfoImplBase &BFI,
483                                      const Scaled64 &Min, const Scaled64 &Max) {
484   // Scale the Factor to a size that creates integers.  Ideally, integers would
485   // be scaled so that Max == UINT64_MAX so that they can be best
486   // differentiated.  However, in the presence of large frequency values, small
487   // frequencies are scaled down to 1, making it impossible to differentiate
488   // small, unequal numbers. When the spread between Min and Max frequencies
489   // fits well within MaxBits, we make the scale be at least 8.
490   const unsigned MaxBits = 64;
491   const unsigned SpreadBits = (Max / Min).lg();
492   Scaled64 ScalingFactor;
493   if (SpreadBits <= MaxBits - 3) {
494     // If the values are small enough, make the scaling factor at least 8 to
495     // allow distinguishing small values.
496     ScalingFactor = Min.inverse();
497     ScalingFactor <<= 3;
498   } else {
499     // If the values need more than MaxBits to be represented, saturate small
500     // frequency values down to 1 by using a scaling factor that benefits large
501     // frequency values.
502     ScalingFactor = Scaled64(1, MaxBits) / Max;
503   }
504 
505   // Translate the floats to integers.
506   LLVM_DEBUG(dbgs() << "float-to-int: min = " << Min << ", max = " << Max
507                     << ", factor = " << ScalingFactor << "\n");
508   for (size_t Index = 0; Index < BFI.Freqs.size(); ++Index) {
509     Scaled64 Scaled = BFI.Freqs[Index].Scaled * ScalingFactor;
510     BFI.Freqs[Index].Integer = std::max(UINT64_C(1), Scaled.toInt<uint64_t>());
511     LLVM_DEBUG(dbgs() << " - " << BFI.getBlockName(Index) << ": float = "
512                       << BFI.Freqs[Index].Scaled << ", scaled = " << Scaled
513                       << ", int = " << BFI.Freqs[Index].Integer << "\n");
514   }
515 }
516 
517 /// Unwrap a loop package.
518 ///
519 /// Visits all the members of a loop, adjusting their BlockData according to
520 /// the loop's pseudo-node.
521 static void unwrapLoop(BlockFrequencyInfoImplBase &BFI, LoopData &Loop) {
522   LLVM_DEBUG(dbgs() << "unwrap-loop-package: " << BFI.getLoopName(Loop)
523                     << ": mass = " << Loop.Mass << ", scale = " << Loop.Scale
524                     << "\n");
525   Loop.Scale *= Loop.Mass.toScaled();
526   Loop.IsPackaged = false;
527   LLVM_DEBUG(dbgs() << "  => combined-scale = " << Loop.Scale << "\n");
528 
529   // Propagate the head scale through the loop.  Since members are visited in
530   // RPO, the head scale will be updated by the loop scale first, and then the
531   // final head scale will be used for updated the rest of the members.
532   for (const BlockNode &N : Loop.Nodes) {
533     const auto &Working = BFI.Working[N.Index];
534     Scaled64 &F = Working.isAPackage() ? Working.getPackagedLoop()->Scale
535                                        : BFI.Freqs[N.Index].Scaled;
536     Scaled64 New = Loop.Scale * F;
537     LLVM_DEBUG(dbgs() << " - " << BFI.getBlockName(N) << ": " << F << " => "
538                       << New << "\n");
539     F = New;
540   }
541 }
542 
543 void BlockFrequencyInfoImplBase::unwrapLoops() {
544   // Set initial frequencies from loop-local masses.
545   for (size_t Index = 0; Index < Working.size(); ++Index)
546     Freqs[Index].Scaled = Working[Index].Mass.toScaled();
547 
548   for (LoopData &Loop : Loops)
549     unwrapLoop(*this, Loop);
550 }
551 
552 void BlockFrequencyInfoImplBase::finalizeMetrics() {
553   // Unwrap loop packages in reverse post-order, tracking min and max
554   // frequencies.
555   auto Min = Scaled64::getLargest();
556   auto Max = Scaled64::getZero();
557   for (size_t Index = 0; Index < Working.size(); ++Index) {
558     // Update min/max scale.
559     Min = std::min(Min, Freqs[Index].Scaled);
560     Max = std::max(Max, Freqs[Index].Scaled);
561   }
562 
563   // Convert to integers.
564   convertFloatingToInteger(*this, Min, Max);
565 
566   // Clean up data structures.
567   cleanup(*this);
568 
569   // Print out the final stats.
570   LLVM_DEBUG(dump());
571 }
572 
573 BlockFrequency
574 BlockFrequencyInfoImplBase::getBlockFreq(const BlockNode &Node) const {
575   if (!Node.isValid()) {
576 #ifndef NDEBUG
577     if (CheckBFIUnknownBlockQueries) {
578       SmallString<256> Msg;
579       raw_svector_ostream OS(Msg);
580       OS << "*** Detected BFI query for unknown block " << getBlockName(Node);
581       report_fatal_error(OS.str());
582     }
583 #endif
584     return 0;
585   }
586   return Freqs[Node.Index].Integer;
587 }
588 
589 Optional<uint64_t>
590 BlockFrequencyInfoImplBase::getBlockProfileCount(const Function &F,
591                                                  const BlockNode &Node,
592                                                  bool AllowSynthetic) const {
593   return getProfileCountFromFreq(F, getBlockFreq(Node).getFrequency(),
594                                  AllowSynthetic);
595 }
596 
597 Optional<uint64_t>
598 BlockFrequencyInfoImplBase::getProfileCountFromFreq(const Function &F,
599                                                     uint64_t Freq,
600                                                     bool AllowSynthetic) const {
601   auto EntryCount = F.getEntryCount(AllowSynthetic);
602   if (!EntryCount)
603     return None;
604   // Use 128 bit APInt to do the arithmetic to avoid overflow.
605   APInt BlockCount(128, EntryCount.getCount());
606   APInt BlockFreq(128, Freq);
607   APInt EntryFreq(128, getEntryFreq());
608   BlockCount *= BlockFreq;
609   // Rounded division of BlockCount by EntryFreq. Since EntryFreq is unsigned
610   // lshr by 1 gives EntryFreq/2.
611   BlockCount = (BlockCount + EntryFreq.lshr(1)).udiv(EntryFreq);
612   return BlockCount.getLimitedValue();
613 }
614 
615 bool
616 BlockFrequencyInfoImplBase::isIrrLoopHeader(const BlockNode &Node) {
617   if (!Node.isValid())
618     return false;
619   return IsIrrLoopHeader.test(Node.Index);
620 }
621 
622 Scaled64
623 BlockFrequencyInfoImplBase::getFloatingBlockFreq(const BlockNode &Node) const {
624   if (!Node.isValid())
625     return Scaled64::getZero();
626   return Freqs[Node.Index].Scaled;
627 }
628 
629 void BlockFrequencyInfoImplBase::setBlockFreq(const BlockNode &Node,
630                                               uint64_t Freq) {
631   assert(Node.isValid() && "Expected valid node");
632   assert(Node.Index < Freqs.size() && "Expected legal index");
633   Freqs[Node.Index].Integer = Freq;
634 }
635 
636 std::string
637 BlockFrequencyInfoImplBase::getBlockName(const BlockNode &Node) const {
638   return {};
639 }
640 
641 std::string
642 BlockFrequencyInfoImplBase::getLoopName(const LoopData &Loop) const {
643   return getBlockName(Loop.getHeader()) + (Loop.isIrreducible() ? "**" : "*");
644 }
645 
646 raw_ostream &
647 BlockFrequencyInfoImplBase::printBlockFreq(raw_ostream &OS,
648                                            const BlockNode &Node) const {
649   return OS << getFloatingBlockFreq(Node);
650 }
651 
652 raw_ostream &
653 BlockFrequencyInfoImplBase::printBlockFreq(raw_ostream &OS,
654                                            const BlockFrequency &Freq) const {
655   Scaled64 Block(Freq.getFrequency(), 0);
656   Scaled64 Entry(getEntryFreq(), 0);
657 
658   return OS << Block / Entry;
659 }
660 
661 void IrreducibleGraph::addNodesInLoop(const BFIBase::LoopData &OuterLoop) {
662   Start = OuterLoop.getHeader();
663   Nodes.reserve(OuterLoop.Nodes.size());
664   for (auto N : OuterLoop.Nodes)
665     addNode(N);
666   indexNodes();
667 }
668 
669 void IrreducibleGraph::addNodesInFunction() {
670   Start = 0;
671   for (uint32_t Index = 0; Index < BFI.Working.size(); ++Index)
672     if (!BFI.Working[Index].isPackaged())
673       addNode(Index);
674   indexNodes();
675 }
676 
677 void IrreducibleGraph::indexNodes() {
678   for (auto &I : Nodes)
679     Lookup[I.Node.Index] = &I;
680 }
681 
682 void IrreducibleGraph::addEdge(IrrNode &Irr, const BlockNode &Succ,
683                                const BFIBase::LoopData *OuterLoop) {
684   if (OuterLoop && OuterLoop->isHeader(Succ))
685     return;
686   auto L = Lookup.find(Succ.Index);
687   if (L == Lookup.end())
688     return;
689   IrrNode &SuccIrr = *L->second;
690   Irr.Edges.push_back(&SuccIrr);
691   SuccIrr.Edges.push_front(&Irr);
692   ++SuccIrr.NumIn;
693 }
694 
695 namespace llvm {
696 
697 template <> struct GraphTraits<IrreducibleGraph> {
698   using GraphT = bfi_detail::IrreducibleGraph;
699   using NodeRef = const GraphT::IrrNode *;
700   using ChildIteratorType = GraphT::IrrNode::iterator;
701 
702   static NodeRef getEntryNode(const GraphT &G) { return G.StartIrr; }
703   static ChildIteratorType child_begin(NodeRef N) { return N->succ_begin(); }
704   static ChildIteratorType child_end(NodeRef N) { return N->succ_end(); }
705 };
706 
707 } // end namespace llvm
708 
709 /// Find extra irreducible headers.
710 ///
711 /// Find entry blocks and other blocks with backedges, which exist when \c G
712 /// contains irreducible sub-SCCs.
713 static void findIrreducibleHeaders(
714     const BlockFrequencyInfoImplBase &BFI,
715     const IrreducibleGraph &G,
716     const std::vector<const IrreducibleGraph::IrrNode *> &SCC,
717     LoopData::NodeList &Headers, LoopData::NodeList &Others) {
718   // Map from nodes in the SCC to whether it's an entry block.
719   SmallDenseMap<const IrreducibleGraph::IrrNode *, bool, 8> InSCC;
720 
721   // InSCC also acts the set of nodes in the graph.  Seed it.
722   for (const auto *I : SCC)
723     InSCC[I] = false;
724 
725   for (auto I = InSCC.begin(), E = InSCC.end(); I != E; ++I) {
726     auto &Irr = *I->first;
727     for (const auto *P : make_range(Irr.pred_begin(), Irr.pred_end())) {
728       if (InSCC.count(P))
729         continue;
730 
731       // This is an entry block.
732       I->second = true;
733       Headers.push_back(Irr.Node);
734       LLVM_DEBUG(dbgs() << "  => entry = " << BFI.getBlockName(Irr.Node)
735                         << "\n");
736       break;
737     }
738   }
739   assert(Headers.size() >= 2 &&
740          "Expected irreducible CFG; -loop-info is likely invalid");
741   if (Headers.size() == InSCC.size()) {
742     // Every block is a header.
743     llvm::sort(Headers);
744     return;
745   }
746 
747   // Look for extra headers from irreducible sub-SCCs.
748   for (const auto &I : InSCC) {
749     // Entry blocks are already headers.
750     if (I.second)
751       continue;
752 
753     auto &Irr = *I.first;
754     for (const auto *P : make_range(Irr.pred_begin(), Irr.pred_end())) {
755       // Skip forward edges.
756       if (P->Node < Irr.Node)
757         continue;
758 
759       // Skip predecessors from entry blocks.  These can have inverted
760       // ordering.
761       if (InSCC.lookup(P))
762         continue;
763 
764       // Store the extra header.
765       Headers.push_back(Irr.Node);
766       LLVM_DEBUG(dbgs() << "  => extra = " << BFI.getBlockName(Irr.Node)
767                         << "\n");
768       break;
769     }
770     if (Headers.back() == Irr.Node)
771       // Added this as a header.
772       continue;
773 
774     // This is not a header.
775     Others.push_back(Irr.Node);
776     LLVM_DEBUG(dbgs() << "  => other = " << BFI.getBlockName(Irr.Node) << "\n");
777   }
778   llvm::sort(Headers);
779   llvm::sort(Others);
780 }
781 
782 static void createIrreducibleLoop(
783     BlockFrequencyInfoImplBase &BFI, const IrreducibleGraph &G,
784     LoopData *OuterLoop, std::list<LoopData>::iterator Insert,
785     const std::vector<const IrreducibleGraph::IrrNode *> &SCC) {
786   // Translate the SCC into RPO.
787   LLVM_DEBUG(dbgs() << " - found-scc\n");
788 
789   LoopData::NodeList Headers;
790   LoopData::NodeList Others;
791   findIrreducibleHeaders(BFI, G, SCC, Headers, Others);
792 
793   auto Loop = BFI.Loops.emplace(Insert, OuterLoop, Headers.begin(),
794                                 Headers.end(), Others.begin(), Others.end());
795 
796   // Update loop hierarchy.
797   for (const auto &N : Loop->Nodes)
798     if (BFI.Working[N.Index].isLoopHeader())
799       BFI.Working[N.Index].Loop->Parent = &*Loop;
800     else
801       BFI.Working[N.Index].Loop = &*Loop;
802 }
803 
804 iterator_range<std::list<LoopData>::iterator>
805 BlockFrequencyInfoImplBase::analyzeIrreducible(
806     const IrreducibleGraph &G, LoopData *OuterLoop,
807     std::list<LoopData>::iterator Insert) {
808   assert((OuterLoop == nullptr) == (Insert == Loops.begin()));
809   auto Prev = OuterLoop ? std::prev(Insert) : Loops.end();
810 
811   for (auto I = scc_begin(G); !I.isAtEnd(); ++I) {
812     if (I->size() < 2)
813       continue;
814 
815     // Translate the SCC into RPO.
816     createIrreducibleLoop(*this, G, OuterLoop, Insert, *I);
817   }
818 
819   if (OuterLoop)
820     return make_range(std::next(Prev), Insert);
821   return make_range(Loops.begin(), Insert);
822 }
823 
824 void
825 BlockFrequencyInfoImplBase::updateLoopWithIrreducible(LoopData &OuterLoop) {
826   OuterLoop.Exits.clear();
827   for (auto &Mass : OuterLoop.BackedgeMass)
828     Mass = BlockMass::getEmpty();
829   auto O = OuterLoop.Nodes.begin() + 1;
830   for (auto I = O, E = OuterLoop.Nodes.end(); I != E; ++I)
831     if (!Working[I->Index].isPackaged())
832       *O++ = *I;
833   OuterLoop.Nodes.erase(O, OuterLoop.Nodes.end());
834 }
835 
836 void BlockFrequencyInfoImplBase::adjustLoopHeaderMass(LoopData &Loop) {
837   assert(Loop.isIrreducible() && "this only makes sense on irreducible loops");
838 
839   // Since the loop has more than one header block, the mass flowing back into
840   // each header will be different. Adjust the mass in each header loop to
841   // reflect the masses flowing through back edges.
842   //
843   // To do this, we distribute the initial mass using the backedge masses
844   // as weights for the distribution.
845   BlockMass LoopMass = BlockMass::getFull();
846   Distribution Dist;
847 
848   LLVM_DEBUG(dbgs() << "adjust-loop-header-mass:\n");
849   for (uint32_t H = 0; H < Loop.NumHeaders; ++H) {
850     auto &HeaderNode = Loop.Nodes[H];
851     auto &BackedgeMass = Loop.BackedgeMass[Loop.getHeaderIndex(HeaderNode)];
852     LLVM_DEBUG(dbgs() << " - Add back edge mass for node "
853                       << getBlockName(HeaderNode) << ": " << BackedgeMass
854                       << "\n");
855     if (BackedgeMass.getMass() > 0)
856       Dist.addLocal(HeaderNode, BackedgeMass.getMass());
857     else
858       LLVM_DEBUG(dbgs() << "   Nothing added. Back edge mass is zero\n");
859   }
860 
861   DitheringDistributer D(Dist, LoopMass);
862 
863   LLVM_DEBUG(dbgs() << " Distribute loop mass " << LoopMass
864                     << " to headers using above weights\n");
865   for (const Weight &W : Dist.Weights) {
866     BlockMass Taken = D.takeMass(W.Amount);
867     assert(W.Type == Weight::Local && "all weights should be local");
868     Working[W.TargetNode.Index].getMass() = Taken;
869     LLVM_DEBUG(debugAssign(*this, D, W.TargetNode, Taken, nullptr));
870   }
871 }
872 
873 void BlockFrequencyInfoImplBase::distributeIrrLoopHeaderMass(Distribution &Dist) {
874   BlockMass LoopMass = BlockMass::getFull();
875   DitheringDistributer D(Dist, LoopMass);
876   for (const Weight &W : Dist.Weights) {
877     BlockMass Taken = D.takeMass(W.Amount);
878     assert(W.Type == Weight::Local && "all weights should be local");
879     Working[W.TargetNode.Index].getMass() = Taken;
880     LLVM_DEBUG(debugAssign(*this, D, W.TargetNode, Taken, nullptr));
881   }
882 }
883