Home | History | Annotate | Download | only in CodeGen
      1 //===-- RegAllocGreedy.cpp - greedy register allocator --------------------===//
      2 //
      3 //                     The LLVM Compiler Infrastructure
      4 //
      5 // This file is distributed under the University of Illinois Open Source
      6 // License. See LICENSE.TXT for details.
      7 //
      8 //===----------------------------------------------------------------------===//
      9 //
     10 // This file defines the RAGreedy function pass for register allocation in
     11 // optimized builds.
     12 //
     13 //===----------------------------------------------------------------------===//
     14 
     15 #define DEBUG_TYPE "regalloc"
     16 #include "llvm/CodeGen/Passes.h"
     17 #include "AllocationOrder.h"
     18 #include "InterferenceCache.h"
     19 #include "LiveDebugVariables.h"
     20 #include "RegAllocBase.h"
     21 #include "SpillPlacement.h"
     22 #include "Spiller.h"
     23 #include "SplitKit.h"
     24 #include "llvm/ADT/Statistic.h"
     25 #include "llvm/Analysis/AliasAnalysis.h"
     26 #include "llvm/CodeGen/CalcSpillWeights.h"
     27 #include "llvm/CodeGen/EdgeBundles.h"
     28 #include "llvm/CodeGen/LiveIntervalAnalysis.h"
     29 #include "llvm/CodeGen/LiveRangeEdit.h"
     30 #include "llvm/CodeGen/LiveRegMatrix.h"
     31 #include "llvm/CodeGen/LiveStackAnalysis.h"
     32 #include "llvm/CodeGen/MachineBlockFrequencyInfo.h"
     33 #include "llvm/CodeGen/MachineDominators.h"
     34 #include "llvm/CodeGen/MachineFunctionPass.h"
     35 #include "llvm/CodeGen/MachineLoopInfo.h"
     36 #include "llvm/CodeGen/MachineRegisterInfo.h"
     37 #include "llvm/CodeGen/RegAllocRegistry.h"
     38 #include "llvm/CodeGen/VirtRegMap.h"
     39 #include "llvm/PassAnalysisSupport.h"
     40 #include "llvm/Support/CommandLine.h"
     41 #include "llvm/Support/Debug.h"
     42 #include "llvm/Support/ErrorHandling.h"
     43 #include "llvm/Support/Timer.h"
     44 #include "llvm/Support/raw_ostream.h"
     45 #include <queue>
     46 
     47 using namespace llvm;
     48 
     49 STATISTIC(NumGlobalSplits, "Number of split global live ranges");
     50 STATISTIC(NumLocalSplits,  "Number of split local live ranges");
     51 STATISTIC(NumEvicted,      "Number of interferences evicted");
     52 
     53 static cl::opt<SplitEditor::ComplementSpillMode>
     54 SplitSpillMode("split-spill-mode", cl::Hidden,
     55   cl::desc("Spill mode for splitting live ranges"),
     56   cl::values(clEnumValN(SplitEditor::SM_Partition, "default", "Default"),
     57              clEnumValN(SplitEditor::SM_Size,  "size",  "Optimize for size"),
     58              clEnumValN(SplitEditor::SM_Speed, "speed", "Optimize for speed"),
     59              clEnumValEnd),
     60   cl::init(SplitEditor::SM_Partition));
     61 
     62 static RegisterRegAlloc greedyRegAlloc("greedy", "greedy register allocator",
     63                                        createGreedyRegisterAllocator);
     64 
     65 namespace {
     66 class RAGreedy : public MachineFunctionPass,
     67                  public RegAllocBase,
     68                  private LiveRangeEdit::Delegate {
     69 
     70   // context
     71   MachineFunction *MF;
     72 
     73   // analyses
     74   SlotIndexes *Indexes;
     75   MachineBlockFrequencyInfo *MBFI;
     76   MachineDominatorTree *DomTree;
     77   MachineLoopInfo *Loops;
     78   EdgeBundles *Bundles;
     79   SpillPlacement *SpillPlacer;
     80   LiveDebugVariables *DebugVars;
     81 
     82   // state
     83   OwningPtr<Spiller> SpillerInstance;
     84   std::priority_queue<std::pair<unsigned, unsigned> > Queue;
     85   unsigned NextCascade;
     86 
     87   // Live ranges pass through a number of stages as we try to allocate them.
     88   // Some of the stages may also create new live ranges:
     89   //
     90   // - Region splitting.
     91   // - Per-block splitting.
     92   // - Local splitting.
     93   // - Spilling.
     94   //
     95   // Ranges produced by one of the stages skip the previous stages when they are
     96   // dequeued. This improves performance because we can skip interference checks
     97   // that are unlikely to give any results. It also guarantees that the live
     98   // range splitting algorithm terminates, something that is otherwise hard to
     99   // ensure.
    100   enum LiveRangeStage {
    101     /// Newly created live range that has never been queued.
    102     RS_New,
    103 
    104     /// Only attempt assignment and eviction. Then requeue as RS_Split.
    105     RS_Assign,
    106 
    107     /// Attempt live range splitting if assignment is impossible.
    108     RS_Split,
    109 
    110     /// Attempt more aggressive live range splitting that is guaranteed to make
    111     /// progress.  This is used for split products that may not be making
    112     /// progress.
    113     RS_Split2,
    114 
    115     /// Live range will be spilled.  No more splitting will be attempted.
    116     RS_Spill,
    117 
    118     /// There is nothing more we can do to this live range.  Abort compilation
    119     /// if it can't be assigned.
    120     RS_Done
    121   };
    122 
    123   static const char *const StageName[];
    124 
    125   // RegInfo - Keep additional information about each live range.
    126   struct RegInfo {
    127     LiveRangeStage Stage;
    128 
    129     // Cascade - Eviction loop prevention. See canEvictInterference().
    130     unsigned Cascade;
    131 
    132     RegInfo() : Stage(RS_New), Cascade(0) {}
    133   };
    134 
    135   IndexedMap<RegInfo, VirtReg2IndexFunctor> ExtraRegInfo;
    136 
    137   LiveRangeStage getStage(const LiveInterval &VirtReg) const {
    138     return ExtraRegInfo[VirtReg.reg].Stage;
    139   }
    140 
    141   void setStage(const LiveInterval &VirtReg, LiveRangeStage Stage) {
    142     ExtraRegInfo.resize(MRI->getNumVirtRegs());
    143     ExtraRegInfo[VirtReg.reg].Stage = Stage;
    144   }
    145 
    146   template<typename Iterator>
    147   void setStage(Iterator Begin, Iterator End, LiveRangeStage NewStage) {
    148     ExtraRegInfo.resize(MRI->getNumVirtRegs());
    149     for (;Begin != End; ++Begin) {
    150       unsigned Reg = (*Begin)->reg;
    151       if (ExtraRegInfo[Reg].Stage == RS_New)
    152         ExtraRegInfo[Reg].Stage = NewStage;
    153     }
    154   }
    155 
    156   /// Cost of evicting interference.
    157   struct EvictionCost {
    158     unsigned BrokenHints; ///< Total number of broken hints.
    159     float MaxWeight;      ///< Maximum spill weight evicted.
    160 
    161     EvictionCost(unsigned B = 0) : BrokenHints(B), MaxWeight(0) {}
    162 
    163     bool isMax() const { return BrokenHints == ~0u; }
    164 
    165     bool operator<(const EvictionCost &O) const {
    166       if (BrokenHints != O.BrokenHints)
    167         return BrokenHints < O.BrokenHints;
    168       return MaxWeight < O.MaxWeight;
    169     }
    170   };
    171 
    172   // splitting state.
    173   OwningPtr<SplitAnalysis> SA;
    174   OwningPtr<SplitEditor> SE;
    175 
    176   /// Cached per-block interference maps
    177   InterferenceCache IntfCache;
    178 
    179   /// All basic blocks where the current register has uses.
    180   SmallVector<SpillPlacement::BlockConstraint, 8> SplitConstraints;
    181 
    182   /// Global live range splitting candidate info.
    183   struct GlobalSplitCandidate {
    184     // Register intended for assignment, or 0.
    185     unsigned PhysReg;
    186 
    187     // SplitKit interval index for this candidate.
    188     unsigned IntvIdx;
    189 
    190     // Interference for PhysReg.
    191     InterferenceCache::Cursor Intf;
    192 
    193     // Bundles where this candidate should be live.
    194     BitVector LiveBundles;
    195     SmallVector<unsigned, 8> ActiveBlocks;
    196 
    197     void reset(InterferenceCache &Cache, unsigned Reg) {
    198       PhysReg = Reg;
    199       IntvIdx = 0;
    200       Intf.setPhysReg(Cache, Reg);
    201       LiveBundles.clear();
    202       ActiveBlocks.clear();
    203     }
    204 
    205     // Set B[i] = C for every live bundle where B[i] was NoCand.
    206     unsigned getBundles(SmallVectorImpl<unsigned> &B, unsigned C) {
    207       unsigned Count = 0;
    208       for (int i = LiveBundles.find_first(); i >= 0;
    209            i = LiveBundles.find_next(i))
    210         if (B[i] == NoCand) {
    211           B[i] = C;
    212           Count++;
    213         }
    214       return Count;
    215     }
    216   };
    217 
    218   /// Candidate info for for each PhysReg in AllocationOrder.
    219   /// This vector never shrinks, but grows to the size of the largest register
    220   /// class.
    221   SmallVector<GlobalSplitCandidate, 32> GlobalCand;
    222 
    223   enum { NoCand = ~0u };
    224 
    225   /// Candidate map. Each edge bundle is assigned to a GlobalCand entry, or to
    226   /// NoCand which indicates the stack interval.
    227   SmallVector<unsigned, 32> BundleCand;
    228 
    229 public:
    230   RAGreedy();
    231 
    232   /// Return the pass name.
    233   virtual const char* getPassName() const {
    234     return "Greedy Register Allocator";
    235   }
    236 
    237   /// RAGreedy analysis usage.
    238   virtual void getAnalysisUsage(AnalysisUsage &AU) const;
    239   virtual void releaseMemory();
    240   virtual Spiller &spiller() { return *SpillerInstance; }
    241   virtual void enqueue(LiveInterval *LI);
    242   virtual LiveInterval *dequeue();
    243   virtual unsigned selectOrSplit(LiveInterval&,
    244                                  SmallVectorImpl<LiveInterval*>&);
    245 
    246   /// Perform register allocation.
    247   virtual bool runOnMachineFunction(MachineFunction &mf);
    248 
    249   static char ID;
    250 
    251 private:
    252   bool LRE_CanEraseVirtReg(unsigned);
    253   void LRE_WillShrinkVirtReg(unsigned);
    254   void LRE_DidCloneVirtReg(unsigned, unsigned);
    255 
    256   BlockFrequency calcSpillCost();
    257   bool addSplitConstraints(InterferenceCache::Cursor, BlockFrequency&);
    258   void addThroughConstraints(InterferenceCache::Cursor, ArrayRef<unsigned>);
    259   void growRegion(GlobalSplitCandidate &Cand);
    260   BlockFrequency calcGlobalSplitCost(GlobalSplitCandidate&);
    261   bool calcCompactRegion(GlobalSplitCandidate&);
    262   void splitAroundRegion(LiveRangeEdit&, ArrayRef<unsigned>);
    263   void calcGapWeights(unsigned, SmallVectorImpl<float>&);
    264   unsigned canReassign(LiveInterval &VirtReg, unsigned PhysReg);
    265   bool shouldEvict(LiveInterval &A, bool, LiveInterval &B, bool);
    266   bool canEvictInterference(LiveInterval&, unsigned, bool, EvictionCost&);
    267   void evictInterference(LiveInterval&, unsigned,
    268                          SmallVectorImpl<LiveInterval*>&);
    269 
    270   unsigned tryAssign(LiveInterval&, AllocationOrder&,
    271                      SmallVectorImpl<LiveInterval*>&);
    272   unsigned tryEvict(LiveInterval&, AllocationOrder&,
    273                     SmallVectorImpl<LiveInterval*>&, unsigned = ~0u);
    274   unsigned tryRegionSplit(LiveInterval&, AllocationOrder&,
    275                           SmallVectorImpl<LiveInterval*>&);
    276   unsigned tryBlockSplit(LiveInterval&, AllocationOrder&,
    277                          SmallVectorImpl<LiveInterval*>&);
    278   unsigned tryInstructionSplit(LiveInterval&, AllocationOrder&,
    279                                SmallVectorImpl<LiveInterval*>&);
    280   unsigned tryLocalSplit(LiveInterval&, AllocationOrder&,
    281     SmallVectorImpl<LiveInterval*>&);
    282   unsigned trySplit(LiveInterval&, AllocationOrder&,
    283                     SmallVectorImpl<LiveInterval*>&);
    284 };
    285 } // end anonymous namespace
    286 
    287 char RAGreedy::ID = 0;
    288 
    289 #ifndef NDEBUG
    290 const char *const RAGreedy::StageName[] = {
    291     "RS_New",
    292     "RS_Assign",
    293     "RS_Split",
    294     "RS_Split2",
    295     "RS_Spill",
    296     "RS_Done"
    297 };
    298 #endif
    299 
    300 // Hysteresis to use when comparing floats.
    301 // This helps stabilize decisions based on float comparisons.
    302 const float Hysteresis = 0.98f;
    303 
    304 
    305 FunctionPass* llvm::createGreedyRegisterAllocator() {
    306   return new RAGreedy();
    307 }
    308 
    309 RAGreedy::RAGreedy(): MachineFunctionPass(ID) {
    310   initializeLiveDebugVariablesPass(*PassRegistry::getPassRegistry());
    311   initializeSlotIndexesPass(*PassRegistry::getPassRegistry());
    312   initializeLiveIntervalsPass(*PassRegistry::getPassRegistry());
    313   initializeSlotIndexesPass(*PassRegistry::getPassRegistry());
    314   initializeRegisterCoalescerPass(*PassRegistry::getPassRegistry());
    315   initializeMachineSchedulerPass(*PassRegistry::getPassRegistry());
    316   initializeCalculateSpillWeightsPass(*PassRegistry::getPassRegistry());
    317   initializeLiveStacksPass(*PassRegistry::getPassRegistry());
    318   initializeMachineDominatorTreePass(*PassRegistry::getPassRegistry());
    319   initializeMachineLoopInfoPass(*PassRegistry::getPassRegistry());
    320   initializeVirtRegMapPass(*PassRegistry::getPassRegistry());
    321   initializeLiveRegMatrixPass(*PassRegistry::getPassRegistry());
    322   initializeEdgeBundlesPass(*PassRegistry::getPassRegistry());
    323   initializeSpillPlacementPass(*PassRegistry::getPassRegistry());
    324 }
    325 
    326 void RAGreedy::getAnalysisUsage(AnalysisUsage &AU) const {
    327   AU.setPreservesCFG();
    328   AU.addRequired<MachineBlockFrequencyInfo>();
    329   AU.addPreserved<MachineBlockFrequencyInfo>();
    330   AU.addRequired<AliasAnalysis>();
    331   AU.addPreserved<AliasAnalysis>();
    332   AU.addRequired<LiveIntervals>();
    333   AU.addPreserved<LiveIntervals>();
    334   AU.addRequired<SlotIndexes>();
    335   AU.addPreserved<SlotIndexes>();
    336   AU.addRequired<LiveDebugVariables>();
    337   AU.addPreserved<LiveDebugVariables>();
    338   AU.addRequired<LiveStacks>();
    339   AU.addPreserved<LiveStacks>();
    340   AU.addRequired<CalculateSpillWeights>();
    341   AU.addRequired<MachineDominatorTree>();
    342   AU.addPreserved<MachineDominatorTree>();
    343   AU.addRequired<MachineLoopInfo>();
    344   AU.addPreserved<MachineLoopInfo>();
    345   AU.addRequired<VirtRegMap>();
    346   AU.addPreserved<VirtRegMap>();
    347   AU.addRequired<LiveRegMatrix>();
    348   AU.addPreserved<LiveRegMatrix>();
    349   AU.addRequired<EdgeBundles>();
    350   AU.addRequired<SpillPlacement>();
    351   MachineFunctionPass::getAnalysisUsage(AU);
    352 }
    353 
    354 
    355 //===----------------------------------------------------------------------===//
    356 //                     LiveRangeEdit delegate methods
    357 //===----------------------------------------------------------------------===//
    358 
    359 bool RAGreedy::LRE_CanEraseVirtReg(unsigned VirtReg) {
    360   if (VRM->hasPhys(VirtReg)) {
    361     Matrix->unassign(LIS->getInterval(VirtReg));
    362     return true;
    363   }
    364   // Unassigned virtreg is probably in the priority queue.
    365   // RegAllocBase will erase it after dequeueing.
    366   return false;
    367 }
    368 
    369 void RAGreedy::LRE_WillShrinkVirtReg(unsigned VirtReg) {
    370   if (!VRM->hasPhys(VirtReg))
    371     return;
    372 
    373   // Register is assigned, put it back on the queue for reassignment.
    374   LiveInterval &LI = LIS->getInterval(VirtReg);
    375   Matrix->unassign(LI);
    376   enqueue(&LI);
    377 }
    378 
    379 void RAGreedy::LRE_DidCloneVirtReg(unsigned New, unsigned Old) {
    380   // Cloning a register we haven't even heard about yet?  Just ignore it.
    381   if (!ExtraRegInfo.inBounds(Old))
    382     return;
    383 
    384   // LRE may clone a virtual register because dead code elimination causes it to
    385   // be split into connected components. The new components are much smaller
    386   // than the original, so they should get a new chance at being assigned.
    387   // same stage as the parent.
    388   ExtraRegInfo[Old].Stage = RS_Assign;
    389   ExtraRegInfo.grow(New);
    390   ExtraRegInfo[New] = ExtraRegInfo[Old];
    391 }
    392 
    393 void RAGreedy::releaseMemory() {
    394   SpillerInstance.reset(0);
    395   ExtraRegInfo.clear();
    396   GlobalCand.clear();
    397 }
    398 
    399 void RAGreedy::enqueue(LiveInterval *LI) {
    400   // Prioritize live ranges by size, assigning larger ranges first.
    401   // The queue holds (size, reg) pairs.
    402   const unsigned Size = LI->getSize();
    403   const unsigned Reg = LI->reg;
    404   assert(TargetRegisterInfo::isVirtualRegister(Reg) &&
    405          "Can only enqueue virtual registers");
    406   unsigned Prio;
    407 
    408   ExtraRegInfo.grow(Reg);
    409   if (ExtraRegInfo[Reg].Stage == RS_New)
    410     ExtraRegInfo[Reg].Stage = RS_Assign;
    411 
    412   if (ExtraRegInfo[Reg].Stage == RS_Split) {
    413     // Unsplit ranges that couldn't be allocated immediately are deferred until
    414     // everything else has been allocated.
    415     Prio = Size;
    416   } else {
    417     if (ExtraRegInfo[Reg].Stage == RS_Assign && !LI->empty() &&
    418         LIS->intervalIsInOneMBB(*LI)) {
    419       // Allocate original local ranges in linear instruction order. Since they
    420       // are singly defined, this produces optimal coloring in the absence of
    421       // global interference and other constraints.
    422       Prio = LI->beginIndex().getInstrDistance(Indexes->getLastIndex());
    423     }
    424     else {
    425       // Allocate global and split ranges in long->short order. Long ranges that
    426       // don't fit should be spilled (or split) ASAP so they don't create
    427       // interference.  Mark a bit to prioritize global above local ranges.
    428       Prio = (1u << 29) + Size;
    429     }
    430     // Mark a higher bit to prioritize global and local above RS_Split.
    431     Prio |= (1u << 31);
    432 
    433     // Boost ranges that have a physical register hint.
    434     if (VRM->hasKnownPreference(Reg))
    435       Prio |= (1u << 30);
    436   }
    437   // The virtual register number is a tie breaker for same-sized ranges.
    438   // Give lower vreg numbers higher priority to assign them first.
    439   Queue.push(std::make_pair(Prio, ~Reg));
    440 }
    441 
    442 LiveInterval *RAGreedy::dequeue() {
    443   if (Queue.empty())
    444     return 0;
    445   LiveInterval *LI = &LIS->getInterval(~Queue.top().second);
    446   Queue.pop();
    447   return LI;
    448 }
    449 
    450 
    451 //===----------------------------------------------------------------------===//
    452 //                            Direct Assignment
    453 //===----------------------------------------------------------------------===//
    454 
    455 /// tryAssign - Try to assign VirtReg to an available register.
    456 unsigned RAGreedy::tryAssign(LiveInterval &VirtReg,
    457                              AllocationOrder &Order,
    458                              SmallVectorImpl<LiveInterval*> &NewVRegs) {
    459   Order.rewind();
    460   unsigned PhysReg;
    461   while ((PhysReg = Order.next()))
    462     if (!Matrix->checkInterference(VirtReg, PhysReg))
    463       break;
    464   if (!PhysReg || Order.isHint())
    465     return PhysReg;
    466 
    467   // PhysReg is available, but there may be a better choice.
    468 
    469   // If we missed a simple hint, try to cheaply evict interference from the
    470   // preferred register.
    471   if (unsigned Hint = MRI->getSimpleHint(VirtReg.reg))
    472     if (Order.isHint(Hint)) {
    473       DEBUG(dbgs() << "missed hint " << PrintReg(Hint, TRI) << '\n');
    474       EvictionCost MaxCost(1);
    475       if (canEvictInterference(VirtReg, Hint, true, MaxCost)) {
    476         evictInterference(VirtReg, Hint, NewVRegs);
    477         return Hint;
    478       }
    479     }
    480 
    481   // Try to evict interference from a cheaper alternative.
    482   unsigned Cost = TRI->getCostPerUse(PhysReg);
    483 
    484   // Most registers have 0 additional cost.
    485   if (!Cost)
    486     return PhysReg;
    487 
    488   DEBUG(dbgs() << PrintReg(PhysReg, TRI) << " is available at cost " << Cost
    489                << '\n');
    490   unsigned CheapReg = tryEvict(VirtReg, Order, NewVRegs, Cost);
    491   return CheapReg ? CheapReg : PhysReg;
    492 }
    493 
    494 
    495 //===----------------------------------------------------------------------===//
    496 //                         Interference eviction
    497 //===----------------------------------------------------------------------===//
    498 
    499 unsigned RAGreedy::canReassign(LiveInterval &VirtReg, unsigned PrevReg) {
    500   AllocationOrder Order(VirtReg.reg, *VRM, RegClassInfo);
    501   unsigned PhysReg;
    502   while ((PhysReg = Order.next())) {
    503     if (PhysReg == PrevReg)
    504       continue;
    505 
    506     MCRegUnitIterator Units(PhysReg, TRI);
    507     for (; Units.isValid(); ++Units) {
    508       // Instantiate a "subquery", not to be confused with the Queries array.
    509       LiveIntervalUnion::Query subQ(&VirtReg, &Matrix->getLiveUnions()[*Units]);
    510       if (subQ.checkInterference())
    511         break;
    512     }
    513     // If no units have interference, break out with the current PhysReg.
    514     if (!Units.isValid())
    515       break;
    516   }
    517   if (PhysReg)
    518     DEBUG(dbgs() << "can reassign: " << VirtReg << " from "
    519           << PrintReg(PrevReg, TRI) << " to " << PrintReg(PhysReg, TRI)
    520           << '\n');
    521   return PhysReg;
    522 }
    523 
    524 /// shouldEvict - determine if A should evict the assigned live range B. The
    525 /// eviction policy defined by this function together with the allocation order
    526 /// defined by enqueue() decides which registers ultimately end up being split
    527 /// and spilled.
    528 ///
    529 /// Cascade numbers are used to prevent infinite loops if this function is a
    530 /// cyclic relation.
    531 ///
    532 /// @param A          The live range to be assigned.
    533 /// @param IsHint     True when A is about to be assigned to its preferred
    534 ///                   register.
    535 /// @param B          The live range to be evicted.
    536 /// @param BreaksHint True when B is already assigned to its preferred register.
    537 bool RAGreedy::shouldEvict(LiveInterval &A, bool IsHint,
    538                            LiveInterval &B, bool BreaksHint) {
    539   bool CanSplit = getStage(B) < RS_Spill;
    540 
    541   // Be fairly aggressive about following hints as long as the evictee can be
    542   // split.
    543   if (CanSplit && IsHint && !BreaksHint)
    544     return true;
    545 
    546   return A.weight > B.weight;
    547 }
    548 
    549 /// canEvictInterference - Return true if all interferences between VirtReg and
    550 /// PhysReg can be evicted.  When OnlyCheap is set, don't do anything
    551 ///
    552 /// @param VirtReg Live range that is about to be assigned.
    553 /// @param PhysReg Desired register for assignment.
    554 /// @param IsHint  True when PhysReg is VirtReg's preferred register.
    555 /// @param MaxCost Only look for cheaper candidates and update with new cost
    556 ///                when returning true.
    557 /// @returns True when interference can be evicted cheaper than MaxCost.
    558 bool RAGreedy::canEvictInterference(LiveInterval &VirtReg, unsigned PhysReg,
    559                                     bool IsHint, EvictionCost &MaxCost) {
    560   // It is only possible to evict virtual register interference.
    561   if (Matrix->checkInterference(VirtReg, PhysReg) > LiveRegMatrix::IK_VirtReg)
    562     return false;
    563 
    564   bool IsLocal = LIS->intervalIsInOneMBB(VirtReg);
    565 
    566   // Find VirtReg's cascade number. This will be unassigned if VirtReg was never
    567   // involved in an eviction before. If a cascade number was assigned, deny
    568   // evicting anything with the same or a newer cascade number. This prevents
    569   // infinite eviction loops.
    570   //
    571   // This works out so a register without a cascade number is allowed to evict
    572   // anything, and it can be evicted by anything.
    573   unsigned Cascade = ExtraRegInfo[VirtReg.reg].Cascade;
    574   if (!Cascade)
    575     Cascade = NextCascade;
    576 
    577   EvictionCost Cost;
    578   for (MCRegUnitIterator Units(PhysReg, TRI); Units.isValid(); ++Units) {
    579     LiveIntervalUnion::Query &Q = Matrix->query(VirtReg, *Units);
    580     // If there is 10 or more interferences, chances are one is heavier.
    581     if (Q.collectInterferingVRegs(10) >= 10)
    582       return false;
    583 
    584     // Check if any interfering live range is heavier than MaxWeight.
    585     for (unsigned i = Q.interferingVRegs().size(); i; --i) {
    586       LiveInterval *Intf = Q.interferingVRegs()[i - 1];
    587       assert(TargetRegisterInfo::isVirtualRegister(Intf->reg) &&
    588              "Only expecting virtual register interference from query");
    589       // Never evict spill products. They cannot split or spill.
    590       if (getStage(*Intf) == RS_Done)
    591         return false;
    592       // Once a live range becomes small enough, it is urgent that we find a
    593       // register for it. This is indicated by an infinite spill weight. These
    594       // urgent live ranges get to evict almost anything.
    595       //
    596       // Also allow urgent evictions of unspillable ranges from a strictly
    597       // larger allocation order.
    598       bool Urgent = !VirtReg.isSpillable() &&
    599         (Intf->isSpillable() ||
    600          RegClassInfo.getNumAllocatableRegs(MRI->getRegClass(VirtReg.reg)) <
    601          RegClassInfo.getNumAllocatableRegs(MRI->getRegClass(Intf->reg)));
    602       // Only evict older cascades or live ranges without a cascade.
    603       unsigned IntfCascade = ExtraRegInfo[Intf->reg].Cascade;
    604       if (Cascade <= IntfCascade) {
    605         if (!Urgent)
    606           return false;
    607         // We permit breaking cascades for urgent evictions. It should be the
    608         // last resort, though, so make it really expensive.
    609         Cost.BrokenHints += 10;
    610       }
    611       // Would this break a satisfied hint?
    612       bool BreaksHint = VRM->hasPreferredPhys(Intf->reg);
    613       // Update eviction cost.
    614       Cost.BrokenHints += BreaksHint;
    615       Cost.MaxWeight = std::max(Cost.MaxWeight, Intf->weight);
    616       // Abort if this would be too expensive.
    617       if (!(Cost < MaxCost))
    618         return false;
    619       if (Urgent)
    620         continue;
    621       // If !MaxCost.isMax(), then we're just looking for a cheap register.
    622       // Evicting another local live range in this case could lead to suboptimal
    623       // coloring.
    624       if (!MaxCost.isMax() && IsLocal && LIS->intervalIsInOneMBB(*Intf) &&
    625           !canReassign(*Intf, PhysReg)) {
    626         return false;
    627       }
    628       // Finally, apply the eviction policy for non-urgent evictions.
    629       if (!shouldEvict(VirtReg, IsHint, *Intf, BreaksHint))
    630         return false;
    631     }
    632   }
    633   MaxCost = Cost;
    634   return true;
    635 }
    636 
    637 /// evictInterference - Evict any interferring registers that prevent VirtReg
    638 /// from being assigned to Physreg. This assumes that canEvictInterference
    639 /// returned true.
    640 void RAGreedy::evictInterference(LiveInterval &VirtReg, unsigned PhysReg,
    641                                  SmallVectorImpl<LiveInterval*> &NewVRegs) {
    642   // Make sure that VirtReg has a cascade number, and assign that cascade
    643   // number to every evicted register. These live ranges than then only be
    644   // evicted by a newer cascade, preventing infinite loops.
    645   unsigned Cascade = ExtraRegInfo[VirtReg.reg].Cascade;
    646   if (!Cascade)
    647     Cascade = ExtraRegInfo[VirtReg.reg].Cascade = NextCascade++;
    648 
    649   DEBUG(dbgs() << "evicting " << PrintReg(PhysReg, TRI)
    650                << " interference: Cascade " << Cascade << '\n');
    651 
    652   // Collect all interfering virtregs first.
    653   SmallVector<LiveInterval*, 8> Intfs;
    654   for (MCRegUnitIterator Units(PhysReg, TRI); Units.isValid(); ++Units) {
    655     LiveIntervalUnion::Query &Q = Matrix->query(VirtReg, *Units);
    656     assert(Q.seenAllInterferences() && "Didn't check all interfererences.");
    657     ArrayRef<LiveInterval*> IVR = Q.interferingVRegs();
    658     Intfs.append(IVR.begin(), IVR.end());
    659   }
    660 
    661   // Evict them second. This will invalidate the queries.
    662   for (unsigned i = 0, e = Intfs.size(); i != e; ++i) {
    663     LiveInterval *Intf = Intfs[i];
    664     // The same VirtReg may be present in multiple RegUnits. Skip duplicates.
    665     if (!VRM->hasPhys(Intf->reg))
    666       continue;
    667     Matrix->unassign(*Intf);
    668     assert((ExtraRegInfo[Intf->reg].Cascade < Cascade ||
    669             VirtReg.isSpillable() < Intf->isSpillable()) &&
    670            "Cannot decrease cascade number, illegal eviction");
    671     ExtraRegInfo[Intf->reg].Cascade = Cascade;
    672     ++NumEvicted;
    673     NewVRegs.push_back(Intf);
    674   }
    675 }
    676 
    677 /// tryEvict - Try to evict all interferences for a physreg.
    678 /// @param  VirtReg Currently unassigned virtual register.
    679 /// @param  Order   Physregs to try.
    680 /// @return         Physreg to assign VirtReg, or 0.
    681 unsigned RAGreedy::tryEvict(LiveInterval &VirtReg,
    682                             AllocationOrder &Order,
    683                             SmallVectorImpl<LiveInterval*> &NewVRegs,
    684                             unsigned CostPerUseLimit) {
    685   NamedRegionTimer T("Evict", TimerGroupName, TimePassesIsEnabled);
    686 
    687   // Keep track of the cheapest interference seen so far.
    688   EvictionCost BestCost(~0u);
    689   unsigned BestPhys = 0;
    690   unsigned OrderLimit = Order.getOrder().size();
    691 
    692   // When we are just looking for a reduced cost per use, don't break any
    693   // hints, and only evict smaller spill weights.
    694   if (CostPerUseLimit < ~0u) {
    695     BestCost.BrokenHints = 0;
    696     BestCost.MaxWeight = VirtReg.weight;
    697 
    698     // Check of any registers in RC are below CostPerUseLimit.
    699     const TargetRegisterClass *RC = MRI->getRegClass(VirtReg.reg);
    700     unsigned MinCost = RegClassInfo.getMinCost(RC);
    701     if (MinCost >= CostPerUseLimit) {
    702       DEBUG(dbgs() << RC->getName() << " minimum cost = " << MinCost
    703                    << ", no cheaper registers to be found.\n");
    704       return 0;
    705     }
    706 
    707     // It is normal for register classes to have a long tail of registers with
    708     // the same cost. We don't need to look at them if they're too expensive.
    709     if (TRI->getCostPerUse(Order.getOrder().back()) >= CostPerUseLimit) {
    710       OrderLimit = RegClassInfo.getLastCostChange(RC);
    711       DEBUG(dbgs() << "Only trying the first " << OrderLimit << " regs.\n");
    712     }
    713   }
    714 
    715   Order.rewind();
    716   while (unsigned PhysReg = Order.nextWithDups(OrderLimit)) {
    717     if (TRI->getCostPerUse(PhysReg) >= CostPerUseLimit)
    718       continue;
    719     // The first use of a callee-saved register in a function has cost 1.
    720     // Don't start using a CSR when the CostPerUseLimit is low.
    721     if (CostPerUseLimit == 1)
    722      if (unsigned CSR = RegClassInfo.getLastCalleeSavedAlias(PhysReg))
    723        if (!MRI->isPhysRegUsed(CSR)) {
    724          DEBUG(dbgs() << PrintReg(PhysReg, TRI) << " would clobber CSR "
    725                       << PrintReg(CSR, TRI) << '\n');
    726          continue;
    727        }
    728 
    729     if (!canEvictInterference(VirtReg, PhysReg, false, BestCost))
    730       continue;
    731 
    732     // Best so far.
    733     BestPhys = PhysReg;
    734 
    735     // Stop if the hint can be used.
    736     if (Order.isHint())
    737       break;
    738   }
    739 
    740   if (!BestPhys)
    741     return 0;
    742 
    743   evictInterference(VirtReg, BestPhys, NewVRegs);
    744   return BestPhys;
    745 }
    746 
    747 
    748 //===----------------------------------------------------------------------===//
    749 //                              Region Splitting
    750 //===----------------------------------------------------------------------===//
    751 
    752 /// addSplitConstraints - Fill out the SplitConstraints vector based on the
    753 /// interference pattern in Physreg and its aliases. Add the constraints to
    754 /// SpillPlacement and return the static cost of this split in Cost, assuming
    755 /// that all preferences in SplitConstraints are met.
    756 /// Return false if there are no bundles with positive bias.
    757 bool RAGreedy::addSplitConstraints(InterferenceCache::Cursor Intf,
    758                                    BlockFrequency &Cost) {
    759   ArrayRef<SplitAnalysis::BlockInfo> UseBlocks = SA->getUseBlocks();
    760 
    761   // Reset interference dependent info.
    762   SplitConstraints.resize(UseBlocks.size());
    763   BlockFrequency StaticCost = 0;
    764   for (unsigned i = 0; i != UseBlocks.size(); ++i) {
    765     const SplitAnalysis::BlockInfo &BI = UseBlocks[i];
    766     SpillPlacement::BlockConstraint &BC = SplitConstraints[i];
    767 
    768     BC.Number = BI.MBB->getNumber();
    769     Intf.moveToBlock(BC.Number);
    770     BC.Entry = BI.LiveIn ? SpillPlacement::PrefReg : SpillPlacement::DontCare;
    771     BC.Exit = BI.LiveOut ? SpillPlacement::PrefReg : SpillPlacement::DontCare;
    772     BC.ChangesValue = BI.FirstDef.isValid();
    773 
    774     if (!Intf.hasInterference())
    775       continue;
    776 
    777     // Number of spill code instructions to insert.
    778     unsigned Ins = 0;
    779 
    780     // Interference for the live-in value.
    781     if (BI.LiveIn) {
    782       if (Intf.first() <= Indexes->getMBBStartIdx(BC.Number))
    783         BC.Entry = SpillPlacement::MustSpill, ++Ins;
    784       else if (Intf.first() < BI.FirstInstr)
    785         BC.Entry = SpillPlacement::PrefSpill, ++Ins;
    786       else if (Intf.first() < BI.LastInstr)
    787         ++Ins;
    788     }
    789 
    790     // Interference for the live-out value.
    791     if (BI.LiveOut) {
    792       if (Intf.last() >= SA->getLastSplitPoint(BC.Number))
    793         BC.Exit = SpillPlacement::MustSpill, ++Ins;
    794       else if (Intf.last() > BI.LastInstr)
    795         BC.Exit = SpillPlacement::PrefSpill, ++Ins;
    796       else if (Intf.last() > BI.FirstInstr)
    797         ++Ins;
    798     }
    799 
    800     // Accumulate the total frequency of inserted spill code.
    801     while (Ins--)
    802       StaticCost += SpillPlacer->getBlockFrequency(BC.Number);
    803   }
    804   Cost = StaticCost;
    805 
    806   // Add constraints for use-blocks. Note that these are the only constraints
    807   // that may add a positive bias, it is downhill from here.
    808   SpillPlacer->addConstraints(SplitConstraints);
    809   return SpillPlacer->scanActiveBundles();
    810 }
    811 
    812 
    813 /// addThroughConstraints - Add constraints and links to SpillPlacer from the
    814 /// live-through blocks in Blocks.
    815 void RAGreedy::addThroughConstraints(InterferenceCache::Cursor Intf,
    816                                      ArrayRef<unsigned> Blocks) {
    817   const unsigned GroupSize = 8;
    818   SpillPlacement::BlockConstraint BCS[GroupSize];
    819   unsigned TBS[GroupSize];
    820   unsigned B = 0, T = 0;
    821 
    822   for (unsigned i = 0; i != Blocks.size(); ++i) {
    823     unsigned Number = Blocks[i];
    824     Intf.moveToBlock(Number);
    825 
    826     if (!Intf.hasInterference()) {
    827       assert(T < GroupSize && "Array overflow");
    828       TBS[T] = Number;
    829       if (++T == GroupSize) {
    830         SpillPlacer->addLinks(makeArrayRef(TBS, T));
    831         T = 0;
    832       }
    833       continue;
    834     }
    835 
    836     assert(B < GroupSize && "Array overflow");
    837     BCS[B].Number = Number;
    838 
    839     // Interference for the live-in value.
    840     if (Intf.first() <= Indexes->getMBBStartIdx(Number))
    841       BCS[B].Entry = SpillPlacement::MustSpill;
    842     else
    843       BCS[B].Entry = SpillPlacement::PrefSpill;
    844 
    845     // Interference for the live-out value.
    846     if (Intf.last() >= SA->getLastSplitPoint(Number))
    847       BCS[B].Exit = SpillPlacement::MustSpill;
    848     else
    849       BCS[B].Exit = SpillPlacement::PrefSpill;
    850 
    851     if (++B == GroupSize) {
    852       ArrayRef<SpillPlacement::BlockConstraint> Array(BCS, B);
    853       SpillPlacer->addConstraints(Array);
    854       B = 0;
    855     }
    856   }
    857 
    858   ArrayRef<SpillPlacement::BlockConstraint> Array(BCS, B);
    859   SpillPlacer->addConstraints(Array);
    860   SpillPlacer->addLinks(makeArrayRef(TBS, T));
    861 }
    862 
    863 void RAGreedy::growRegion(GlobalSplitCandidate &Cand) {
    864   // Keep track of through blocks that have not been added to SpillPlacer.
    865   BitVector Todo = SA->getThroughBlocks();
    866   SmallVectorImpl<unsigned> &ActiveBlocks = Cand.ActiveBlocks;
    867   unsigned AddedTo = 0;
    868 #ifndef NDEBUG
    869   unsigned Visited = 0;
    870 #endif
    871 
    872   for (;;) {
    873     ArrayRef<unsigned> NewBundles = SpillPlacer->getRecentPositive();
    874     // Find new through blocks in the periphery of PrefRegBundles.
    875     for (int i = 0, e = NewBundles.size(); i != e; ++i) {
    876       unsigned Bundle = NewBundles[i];
    877       // Look at all blocks connected to Bundle in the full graph.
    878       ArrayRef<unsigned> Blocks = Bundles->getBlocks(Bundle);
    879       for (ArrayRef<unsigned>::iterator I = Blocks.begin(), E = Blocks.end();
    880            I != E; ++I) {
    881         unsigned Block = *I;
    882         if (!Todo.test(Block))
    883           continue;
    884         Todo.reset(Block);
    885         // This is a new through block. Add it to SpillPlacer later.
    886         ActiveBlocks.push_back(Block);
    887 #ifndef NDEBUG
    888         ++Visited;
    889 #endif
    890       }
    891     }
    892     // Any new blocks to add?
    893     if (ActiveBlocks.size() == AddedTo)
    894       break;
    895 
    896     // Compute through constraints from the interference, or assume that all
    897     // through blocks prefer spilling when forming compact regions.
    898     ArrayRef<unsigned> NewBlocks = makeArrayRef(ActiveBlocks).slice(AddedTo);
    899     if (Cand.PhysReg)
    900       addThroughConstraints(Cand.Intf, NewBlocks);
    901     else
    902       // Provide a strong negative bias on through blocks to prevent unwanted
    903       // liveness on loop backedges.
    904       SpillPlacer->addPrefSpill(NewBlocks, /* Strong= */ true);
    905     AddedTo = ActiveBlocks.size();
    906 
    907     // Perhaps iterating can enable more bundles?
    908     SpillPlacer->iterate();
    909   }
    910   DEBUG(dbgs() << ", v=" << Visited);
    911 }
    912 
    913 /// calcCompactRegion - Compute the set of edge bundles that should be live
    914 /// when splitting the current live range into compact regions.  Compact
    915 /// regions can be computed without looking at interference.  They are the
    916 /// regions formed by removing all the live-through blocks from the live range.
    917 ///
    918 /// Returns false if the current live range is already compact, or if the
    919 /// compact regions would form single block regions anyway.
    920 bool RAGreedy::calcCompactRegion(GlobalSplitCandidate &Cand) {
    921   // Without any through blocks, the live range is already compact.
    922   if (!SA->getNumThroughBlocks())
    923     return false;
    924 
    925   // Compact regions don't correspond to any physreg.
    926   Cand.reset(IntfCache, 0);
    927 
    928   DEBUG(dbgs() << "Compact region bundles");
    929 
    930   // Use the spill placer to determine the live bundles. GrowRegion pretends
    931   // that all the through blocks have interference when PhysReg is unset.
    932   SpillPlacer->prepare(Cand.LiveBundles);
    933 
    934   // The static split cost will be zero since Cand.Intf reports no interference.
    935   BlockFrequency Cost;
    936   if (!addSplitConstraints(Cand.Intf, Cost)) {
    937     DEBUG(dbgs() << ", none.\n");
    938     return false;
    939   }
    940 
    941   growRegion(Cand);
    942   SpillPlacer->finish();
    943 
    944   if (!Cand.LiveBundles.any()) {
    945     DEBUG(dbgs() << ", none.\n");
    946     return false;
    947   }
    948 
    949   DEBUG({
    950     for (int i = Cand.LiveBundles.find_first(); i>=0;
    951          i = Cand.LiveBundles.find_next(i))
    952     dbgs() << " EB#" << i;
    953     dbgs() << ".\n";
    954   });
    955   return true;
    956 }
    957 
    958 /// calcSpillCost - Compute how expensive it would be to split the live range in
    959 /// SA around all use blocks instead of forming bundle regions.
    960 BlockFrequency RAGreedy::calcSpillCost() {
    961   BlockFrequency Cost = 0;
    962   ArrayRef<SplitAnalysis::BlockInfo> UseBlocks = SA->getUseBlocks();
    963   for (unsigned i = 0; i != UseBlocks.size(); ++i) {
    964     const SplitAnalysis::BlockInfo &BI = UseBlocks[i];
    965     unsigned Number = BI.MBB->getNumber();
    966     // We normally only need one spill instruction - a load or a store.
    967     Cost += SpillPlacer->getBlockFrequency(Number);
    968 
    969     // Unless the value is redefined in the block.
    970     if (BI.LiveIn && BI.LiveOut && BI.FirstDef)
    971       Cost += SpillPlacer->getBlockFrequency(Number);
    972   }
    973   return Cost;
    974 }
    975 
    976 /// calcGlobalSplitCost - Return the global split cost of following the split
    977 /// pattern in LiveBundles. This cost should be added to the local cost of the
    978 /// interference pattern in SplitConstraints.
    979 ///
    980 BlockFrequency RAGreedy::calcGlobalSplitCost(GlobalSplitCandidate &Cand) {
    981   BlockFrequency GlobalCost = 0;
    982   const BitVector &LiveBundles = Cand.LiveBundles;
    983   ArrayRef<SplitAnalysis::BlockInfo> UseBlocks = SA->getUseBlocks();
    984   for (unsigned i = 0; i != UseBlocks.size(); ++i) {
    985     const SplitAnalysis::BlockInfo &BI = UseBlocks[i];
    986     SpillPlacement::BlockConstraint &BC = SplitConstraints[i];
    987     bool RegIn  = LiveBundles[Bundles->getBundle(BC.Number, 0)];
    988     bool RegOut = LiveBundles[Bundles->getBundle(BC.Number, 1)];
    989     unsigned Ins = 0;
    990 
    991     if (BI.LiveIn)
    992       Ins += RegIn != (BC.Entry == SpillPlacement::PrefReg);
    993     if (BI.LiveOut)
    994       Ins += RegOut != (BC.Exit == SpillPlacement::PrefReg);
    995     while (Ins--)
    996       GlobalCost += SpillPlacer->getBlockFrequency(BC.Number);
    997   }
    998 
    999   for (unsigned i = 0, e = Cand.ActiveBlocks.size(); i != e; ++i) {
   1000     unsigned Number = Cand.ActiveBlocks[i];
   1001     bool RegIn  = LiveBundles[Bundles->getBundle(Number, 0)];
   1002     bool RegOut = LiveBundles[Bundles->getBundle(Number, 1)];
   1003     if (!RegIn && !RegOut)
   1004       continue;
   1005     if (RegIn && RegOut) {
   1006       // We need double spill code if this block has interference.
   1007       Cand.Intf.moveToBlock(Number);
   1008       if (Cand.Intf.hasInterference()) {
   1009         GlobalCost += SpillPlacer->getBlockFrequency(Number);
   1010         GlobalCost += SpillPlacer->getBlockFrequency(Number);
   1011       }
   1012       continue;
   1013     }
   1014     // live-in / stack-out or stack-in live-out.
   1015     GlobalCost += SpillPlacer->getBlockFrequency(Number);
   1016   }
   1017   return GlobalCost;
   1018 }
   1019 
   1020 /// splitAroundRegion - Split the current live range around the regions
   1021 /// determined by BundleCand and GlobalCand.
   1022 ///
   1023 /// Before calling this function, GlobalCand and BundleCand must be initialized
   1024 /// so each bundle is assigned to a valid candidate, or NoCand for the
   1025 /// stack-bound bundles.  The shared SA/SE SplitAnalysis and SplitEditor
   1026 /// objects must be initialized for the current live range, and intervals
   1027 /// created for the used candidates.
   1028 ///
   1029 /// @param LREdit    The LiveRangeEdit object handling the current split.
   1030 /// @param UsedCands List of used GlobalCand entries. Every BundleCand value
   1031 ///                  must appear in this list.
   1032 void RAGreedy::splitAroundRegion(LiveRangeEdit &LREdit,
   1033                                  ArrayRef<unsigned> UsedCands) {
   1034   // These are the intervals created for new global ranges. We may create more
   1035   // intervals for local ranges.
   1036   const unsigned NumGlobalIntvs = LREdit.size();
   1037   DEBUG(dbgs() << "splitAroundRegion with " << NumGlobalIntvs << " globals.\n");
   1038   assert(NumGlobalIntvs && "No global intervals configured");
   1039 
   1040   // Isolate even single instructions when dealing with a proper sub-class.
   1041   // That guarantees register class inflation for the stack interval because it
   1042   // is all copies.
   1043   unsigned Reg = SA->getParent().reg;
   1044   bool SingleInstrs = RegClassInfo.isProperSubClass(MRI->getRegClass(Reg));
   1045 
   1046   // First handle all the blocks with uses.
   1047   ArrayRef<SplitAnalysis::BlockInfo> UseBlocks = SA->getUseBlocks();
   1048   for (unsigned i = 0; i != UseBlocks.size(); ++i) {
   1049     const SplitAnalysis::BlockInfo &BI = UseBlocks[i];
   1050     unsigned Number = BI.MBB->getNumber();
   1051     unsigned IntvIn = 0, IntvOut = 0;
   1052     SlotIndex IntfIn, IntfOut;
   1053     if (BI.LiveIn) {
   1054       unsigned CandIn = BundleCand[Bundles->getBundle(Number, 0)];
   1055       if (CandIn != NoCand) {
   1056         GlobalSplitCandidate &Cand = GlobalCand[CandIn];
   1057         IntvIn = Cand.IntvIdx;
   1058         Cand.Intf.moveToBlock(Number);
   1059         IntfIn = Cand.Intf.first();
   1060       }
   1061     }
   1062     if (BI.LiveOut) {
   1063       unsigned CandOut = BundleCand[Bundles->getBundle(Number, 1)];
   1064       if (CandOut != NoCand) {
   1065         GlobalSplitCandidate &Cand = GlobalCand[CandOut];
   1066         IntvOut = Cand.IntvIdx;
   1067         Cand.Intf.moveToBlock(Number);
   1068         IntfOut = Cand.Intf.last();
   1069       }
   1070     }
   1071 
   1072     // Create separate intervals for isolated blocks with multiple uses.
   1073     if (!IntvIn && !IntvOut) {
   1074       DEBUG(dbgs() << "BB#" << BI.MBB->getNumber() << " isolated.\n");
   1075       if (SA->shouldSplitSingleBlock(BI, SingleInstrs))
   1076         SE->splitSingleBlock(BI);
   1077       continue;
   1078     }
   1079 
   1080     if (IntvIn && IntvOut)
   1081       SE->splitLiveThroughBlock(Number, IntvIn, IntfIn, IntvOut, IntfOut);
   1082     else if (IntvIn)
   1083       SE->splitRegInBlock(BI, IntvIn, IntfIn);
   1084     else
   1085       SE->splitRegOutBlock(BI, IntvOut, IntfOut);
   1086   }
   1087 
   1088   // Handle live-through blocks. The relevant live-through blocks are stored in
   1089   // the ActiveBlocks list with each candidate. We need to filter out
   1090   // duplicates.
   1091   BitVector Todo = SA->getThroughBlocks();
   1092   for (unsigned c = 0; c != UsedCands.size(); ++c) {
   1093     ArrayRef<unsigned> Blocks = GlobalCand[UsedCands[c]].ActiveBlocks;
   1094     for (unsigned i = 0, e = Blocks.size(); i != e; ++i) {
   1095       unsigned Number = Blocks[i];
   1096       if (!Todo.test(Number))
   1097         continue;
   1098       Todo.reset(Number);
   1099 
   1100       unsigned IntvIn = 0, IntvOut = 0;
   1101       SlotIndex IntfIn, IntfOut;
   1102 
   1103       unsigned CandIn = BundleCand[Bundles->getBundle(Number, 0)];
   1104       if (CandIn != NoCand) {
   1105         GlobalSplitCandidate &Cand = GlobalCand[CandIn];
   1106         IntvIn = Cand.IntvIdx;
   1107         Cand.Intf.moveToBlock(Number);
   1108         IntfIn = Cand.Intf.first();
   1109       }
   1110 
   1111       unsigned CandOut = BundleCand[Bundles->getBundle(Number, 1)];
   1112       if (CandOut != NoCand) {
   1113         GlobalSplitCandidate &Cand = GlobalCand[CandOut];
   1114         IntvOut = Cand.IntvIdx;
   1115         Cand.Intf.moveToBlock(Number);
   1116         IntfOut = Cand.Intf.last();
   1117       }
   1118       if (!IntvIn && !IntvOut)
   1119         continue;
   1120       SE->splitLiveThroughBlock(Number, IntvIn, IntfIn, IntvOut, IntfOut);
   1121     }
   1122   }
   1123 
   1124   ++NumGlobalSplits;
   1125 
   1126   SmallVector<unsigned, 8> IntvMap;
   1127   SE->finish(&IntvMap);
   1128   DebugVars->splitRegister(Reg, LREdit.regs());
   1129 
   1130   ExtraRegInfo.resize(MRI->getNumVirtRegs());
   1131   unsigned OrigBlocks = SA->getNumLiveBlocks();
   1132 
   1133   // Sort out the new intervals created by splitting. We get four kinds:
   1134   // - Remainder intervals should not be split again.
   1135   // - Candidate intervals can be assigned to Cand.PhysReg.
   1136   // - Block-local splits are candidates for local splitting.
   1137   // - DCE leftovers should go back on the queue.
   1138   for (unsigned i = 0, e = LREdit.size(); i != e; ++i) {
   1139     LiveInterval &Reg = *LREdit.get(i);
   1140 
   1141     // Ignore old intervals from DCE.
   1142     if (getStage(Reg) != RS_New)
   1143       continue;
   1144 
   1145     // Remainder interval. Don't try splitting again, spill if it doesn't
   1146     // allocate.
   1147     if (IntvMap[i] == 0) {
   1148       setStage(Reg, RS_Spill);
   1149       continue;
   1150     }
   1151 
   1152     // Global intervals. Allow repeated splitting as long as the number of live
   1153     // blocks is strictly decreasing.
   1154     if (IntvMap[i] < NumGlobalIntvs) {
   1155       if (SA->countLiveBlocks(&Reg) >= OrigBlocks) {
   1156         DEBUG(dbgs() << "Main interval covers the same " << OrigBlocks
   1157                      << " blocks as original.\n");
   1158         // Don't allow repeated splitting as a safe guard against looping.
   1159         setStage(Reg, RS_Split2);
   1160       }
   1161       continue;
   1162     }
   1163 
   1164     // Other intervals are treated as new. This includes local intervals created
   1165     // for blocks with multiple uses, and anything created by DCE.
   1166   }
   1167 
   1168   if (VerifyEnabled)
   1169     MF->verify(this, "After splitting live range around region");
   1170 }
   1171 
   1172 unsigned RAGreedy::tryRegionSplit(LiveInterval &VirtReg, AllocationOrder &Order,
   1173                                   SmallVectorImpl<LiveInterval*> &NewVRegs) {
   1174   unsigned NumCands = 0;
   1175   unsigned BestCand = NoCand;
   1176   BlockFrequency BestCost;
   1177   SmallVector<unsigned, 8> UsedCands;
   1178 
   1179   // Check if we can split this live range around a compact region.
   1180   bool HasCompact = calcCompactRegion(GlobalCand.front());
   1181   if (HasCompact) {
   1182     // Yes, keep GlobalCand[0] as the compact region candidate.
   1183     NumCands = 1;
   1184     BestCost = BlockFrequency::getMaxFrequency();
   1185   } else {
   1186     // No benefit from the compact region, our fallback will be per-block
   1187     // splitting. Make sure we find a solution that is cheaper than spilling.
   1188     BestCost = calcSpillCost();
   1189     DEBUG(dbgs() << "Cost of isolating all blocks = " << BestCost << '\n');
   1190   }
   1191 
   1192   Order.rewind();
   1193   while (unsigned PhysReg = Order.next()) {
   1194     // Discard bad candidates before we run out of interference cache cursors.
   1195     // This will only affect register classes with a lot of registers (>32).
   1196     if (NumCands == IntfCache.getMaxCursors()) {
   1197       unsigned WorstCount = ~0u;
   1198       unsigned Worst = 0;
   1199       for (unsigned i = 0; i != NumCands; ++i) {
   1200         if (i == BestCand || !GlobalCand[i].PhysReg)
   1201           continue;
   1202         unsigned Count = GlobalCand[i].LiveBundles.count();
   1203         if (Count < WorstCount)
   1204           Worst = i, WorstCount = Count;
   1205       }
   1206       --NumCands;
   1207       GlobalCand[Worst] = GlobalCand[NumCands];
   1208       if (BestCand == NumCands)
   1209         BestCand = Worst;
   1210     }
   1211 
   1212     if (GlobalCand.size() <= NumCands)
   1213       GlobalCand.resize(NumCands+1);
   1214     GlobalSplitCandidate &Cand = GlobalCand[NumCands];
   1215     Cand.reset(IntfCache, PhysReg);
   1216 
   1217     SpillPlacer->prepare(Cand.LiveBundles);
   1218     BlockFrequency Cost;
   1219     if (!addSplitConstraints(Cand.Intf, Cost)) {
   1220       DEBUG(dbgs() << PrintReg(PhysReg, TRI) << "\tno positive bundles\n");
   1221       continue;
   1222     }
   1223     DEBUG(dbgs() << PrintReg(PhysReg, TRI) << "\tstatic = " << Cost);
   1224     if (Cost >= BestCost) {
   1225       DEBUG({
   1226         if (BestCand == NoCand)
   1227           dbgs() << " worse than no bundles\n";
   1228         else
   1229           dbgs() << " worse than "
   1230                  << PrintReg(GlobalCand[BestCand].PhysReg, TRI) << '\n';
   1231       });
   1232       continue;
   1233     }
   1234     growRegion(Cand);
   1235 
   1236     SpillPlacer->finish();
   1237 
   1238     // No live bundles, defer to splitSingleBlocks().
   1239     if (!Cand.LiveBundles.any()) {
   1240       DEBUG(dbgs() << " no bundles.\n");
   1241       continue;
   1242     }
   1243 
   1244     Cost += calcGlobalSplitCost(Cand);
   1245     DEBUG({
   1246       dbgs() << ", total = " << Cost << " with bundles";
   1247       for (int i = Cand.LiveBundles.find_first(); i>=0;
   1248            i = Cand.LiveBundles.find_next(i))
   1249         dbgs() << " EB#" << i;
   1250       dbgs() << ".\n";
   1251     });
   1252     if (Cost < BestCost) {
   1253       BestCand = NumCands;
   1254       BestCost = Cost;
   1255     }
   1256     ++NumCands;
   1257   }
   1258 
   1259   // No solutions found, fall back to single block splitting.
   1260   if (!HasCompact && BestCand == NoCand)
   1261     return 0;
   1262 
   1263   // Prepare split editor.
   1264   LiveRangeEdit LREdit(&VirtReg, NewVRegs, *MF, *LIS, VRM, this);
   1265   SE->reset(LREdit, SplitSpillMode);
   1266 
   1267   // Assign all edge bundles to the preferred candidate, or NoCand.
   1268   BundleCand.assign(Bundles->getNumBundles(), NoCand);
   1269 
   1270   // Assign bundles for the best candidate region.
   1271   if (BestCand != NoCand) {
   1272     GlobalSplitCandidate &Cand = GlobalCand[BestCand];
   1273     if (unsigned B = Cand.getBundles(BundleCand, BestCand)) {
   1274       UsedCands.push_back(BestCand);
   1275       Cand.IntvIdx = SE->openIntv();
   1276       DEBUG(dbgs() << "Split for " << PrintReg(Cand.PhysReg, TRI) << " in "
   1277                    << B << " bundles, intv " << Cand.IntvIdx << ".\n");
   1278       (void)B;
   1279     }
   1280   }
   1281 
   1282   // Assign bundles for the compact region.
   1283   if (HasCompact) {
   1284     GlobalSplitCandidate &Cand = GlobalCand.front();
   1285     assert(!Cand.PhysReg && "Compact region has no physreg");
   1286     if (unsigned B = Cand.getBundles(BundleCand, 0)) {
   1287       UsedCands.push_back(0);
   1288       Cand.IntvIdx = SE->openIntv();
   1289       DEBUG(dbgs() << "Split for compact region in " << B << " bundles, intv "
   1290                    << Cand.IntvIdx << ".\n");
   1291       (void)B;
   1292     }
   1293   }
   1294 
   1295   splitAroundRegion(LREdit, UsedCands);
   1296   return 0;
   1297 }
   1298 
   1299 
   1300 //===----------------------------------------------------------------------===//
   1301 //                            Per-Block Splitting
   1302 //===----------------------------------------------------------------------===//
   1303 
   1304 /// tryBlockSplit - Split a global live range around every block with uses. This
   1305 /// creates a lot of local live ranges, that will be split by tryLocalSplit if
   1306 /// they don't allocate.
   1307 unsigned RAGreedy::tryBlockSplit(LiveInterval &VirtReg, AllocationOrder &Order,
   1308                                  SmallVectorImpl<LiveInterval*> &NewVRegs) {
   1309   assert(&SA->getParent() == &VirtReg && "Live range wasn't analyzed");
   1310   unsigned Reg = VirtReg.reg;
   1311   bool SingleInstrs = RegClassInfo.isProperSubClass(MRI->getRegClass(Reg));
   1312   LiveRangeEdit LREdit(&VirtReg, NewVRegs, *MF, *LIS, VRM, this);
   1313   SE->reset(LREdit, SplitSpillMode);
   1314   ArrayRef<SplitAnalysis::BlockInfo> UseBlocks = SA->getUseBlocks();
   1315   for (unsigned i = 0; i != UseBlocks.size(); ++i) {
   1316     const SplitAnalysis::BlockInfo &BI = UseBlocks[i];
   1317     if (SA->shouldSplitSingleBlock(BI, SingleInstrs))
   1318       SE->splitSingleBlock(BI);
   1319   }
   1320   // No blocks were split.
   1321   if (LREdit.empty())
   1322     return 0;
   1323 
   1324   // We did split for some blocks.
   1325   SmallVector<unsigned, 8> IntvMap;
   1326   SE->finish(&IntvMap);
   1327 
   1328   // Tell LiveDebugVariables about the new ranges.
   1329   DebugVars->splitRegister(Reg, LREdit.regs());
   1330 
   1331   ExtraRegInfo.resize(MRI->getNumVirtRegs());
   1332 
   1333   // Sort out the new intervals created by splitting. The remainder interval
   1334   // goes straight to spilling, the new local ranges get to stay RS_New.
   1335   for (unsigned i = 0, e = LREdit.size(); i != e; ++i) {
   1336     LiveInterval &LI = *LREdit.get(i);
   1337     if (getStage(LI) == RS_New && IntvMap[i] == 0)
   1338       setStage(LI, RS_Spill);
   1339   }
   1340 
   1341   if (VerifyEnabled)
   1342     MF->verify(this, "After splitting live range around basic blocks");
   1343   return 0;
   1344 }
   1345 
   1346 
   1347 //===----------------------------------------------------------------------===//
   1348 //                         Per-Instruction Splitting
   1349 //===----------------------------------------------------------------------===//
   1350 
   1351 /// tryInstructionSplit - Split a live range around individual instructions.
   1352 /// This is normally not worthwhile since the spiller is doing essentially the
   1353 /// same thing. However, when the live range is in a constrained register
   1354 /// class, it may help to insert copies such that parts of the live range can
   1355 /// be moved to a larger register class.
   1356 ///
   1357 /// This is similar to spilling to a larger register class.
   1358 unsigned
   1359 RAGreedy::tryInstructionSplit(LiveInterval &VirtReg, AllocationOrder &Order,
   1360                               SmallVectorImpl<LiveInterval*> &NewVRegs) {
   1361   // There is no point to this if there are no larger sub-classes.
   1362   if (!RegClassInfo.isProperSubClass(MRI->getRegClass(VirtReg.reg)))
   1363     return 0;
   1364 
   1365   // Always enable split spill mode, since we're effectively spilling to a
   1366   // register.
   1367   LiveRangeEdit LREdit(&VirtReg, NewVRegs, *MF, *LIS, VRM, this);
   1368   SE->reset(LREdit, SplitEditor::SM_Size);
   1369 
   1370   ArrayRef<SlotIndex> Uses = SA->getUseSlots();
   1371   if (Uses.size() <= 1)
   1372     return 0;
   1373 
   1374   DEBUG(dbgs() << "Split around " << Uses.size() << " individual instrs.\n");
   1375 
   1376   // Split around every non-copy instruction.
   1377   for (unsigned i = 0; i != Uses.size(); ++i) {
   1378     if (const MachineInstr *MI = Indexes->getInstructionFromIndex(Uses[i]))
   1379       if (MI->isFullCopy()) {
   1380         DEBUG(dbgs() << "    skip:\t" << Uses[i] << '\t' << *MI);
   1381         continue;
   1382       }
   1383     SE->openIntv();
   1384     SlotIndex SegStart = SE->enterIntvBefore(Uses[i]);
   1385     SlotIndex SegStop  = SE->leaveIntvAfter(Uses[i]);
   1386     SE->useIntv(SegStart, SegStop);
   1387   }
   1388 
   1389   if (LREdit.empty()) {
   1390     DEBUG(dbgs() << "All uses were copies.\n");
   1391     return 0;
   1392   }
   1393 
   1394   SmallVector<unsigned, 8> IntvMap;
   1395   SE->finish(&IntvMap);
   1396   DebugVars->splitRegister(VirtReg.reg, LREdit.regs());
   1397   ExtraRegInfo.resize(MRI->getNumVirtRegs());
   1398 
   1399   // Assign all new registers to RS_Spill. This was the last chance.
   1400   setStage(LREdit.begin(), LREdit.end(), RS_Spill);
   1401   return 0;
   1402 }
   1403 
   1404 
   1405 //===----------------------------------------------------------------------===//
   1406 //                             Local Splitting
   1407 //===----------------------------------------------------------------------===//
   1408 
   1409 
   1410 /// calcGapWeights - Compute the maximum spill weight that needs to be evicted
   1411 /// in order to use PhysReg between two entries in SA->UseSlots.
   1412 ///
   1413 /// GapWeight[i] represents the gap between UseSlots[i] and UseSlots[i+1].
   1414 ///
   1415 void RAGreedy::calcGapWeights(unsigned PhysReg,
   1416                               SmallVectorImpl<float> &GapWeight) {
   1417   assert(SA->getUseBlocks().size() == 1 && "Not a local interval");
   1418   const SplitAnalysis::BlockInfo &BI = SA->getUseBlocks().front();
   1419   ArrayRef<SlotIndex> Uses = SA->getUseSlots();
   1420   const unsigned NumGaps = Uses.size()-1;
   1421 
   1422   // Start and end points for the interference check.
   1423   SlotIndex StartIdx =
   1424     BI.LiveIn ? BI.FirstInstr.getBaseIndex() : BI.FirstInstr;
   1425   SlotIndex StopIdx =
   1426     BI.LiveOut ? BI.LastInstr.getBoundaryIndex() : BI.LastInstr;
   1427 
   1428   GapWeight.assign(NumGaps, 0.0f);
   1429 
   1430   // Add interference from each overlapping register.
   1431   for (MCRegUnitIterator Units(PhysReg, TRI); Units.isValid(); ++Units) {
   1432     if (!Matrix->query(const_cast<LiveInterval&>(SA->getParent()), *Units)
   1433           .checkInterference())
   1434       continue;
   1435 
   1436     // We know that VirtReg is a continuous interval from FirstInstr to
   1437     // LastInstr, so we don't need InterferenceQuery.
   1438     //
   1439     // Interference that overlaps an instruction is counted in both gaps
   1440     // surrounding the instruction. The exception is interference before
   1441     // StartIdx and after StopIdx.
   1442     //
   1443     LiveIntervalUnion::SegmentIter IntI =
   1444       Matrix->getLiveUnions()[*Units] .find(StartIdx);
   1445     for (unsigned Gap = 0; IntI.valid() && IntI.start() < StopIdx; ++IntI) {
   1446       // Skip the gaps before IntI.
   1447       while (Uses[Gap+1].getBoundaryIndex() < IntI.start())
   1448         if (++Gap == NumGaps)
   1449           break;
   1450       if (Gap == NumGaps)
   1451         break;
   1452 
   1453       // Update the gaps covered by IntI.
   1454       const float weight = IntI.value()->weight;
   1455       for (; Gap != NumGaps; ++Gap) {
   1456         GapWeight[Gap] = std::max(GapWeight[Gap], weight);
   1457         if (Uses[Gap+1].getBaseIndex() >= IntI.stop())
   1458           break;
   1459       }
   1460       if (Gap == NumGaps)
   1461         break;
   1462     }
   1463   }
   1464 
   1465   // Add fixed interference.
   1466   for (MCRegUnitIterator Units(PhysReg, TRI); Units.isValid(); ++Units) {
   1467     const LiveInterval &LI = LIS->getRegUnit(*Units);
   1468     LiveInterval::const_iterator I = LI.find(StartIdx);
   1469     LiveInterval::const_iterator E = LI.end();
   1470 
   1471     // Same loop as above. Mark any overlapped gaps as HUGE_VALF.
   1472     for (unsigned Gap = 0; I != E && I->start < StopIdx; ++I) {
   1473       while (Uses[Gap+1].getBoundaryIndex() < I->start)
   1474         if (++Gap == NumGaps)
   1475           break;
   1476       if (Gap == NumGaps)
   1477         break;
   1478 
   1479       for (; Gap != NumGaps; ++Gap) {
   1480         GapWeight[Gap] = HUGE_VALF;
   1481         if (Uses[Gap+1].getBaseIndex() >= I->end)
   1482           break;
   1483       }
   1484       if (Gap == NumGaps)
   1485         break;
   1486     }
   1487   }
   1488 }
   1489 
   1490 /// tryLocalSplit - Try to split VirtReg into smaller intervals inside its only
   1491 /// basic block.
   1492 ///
   1493 unsigned RAGreedy::tryLocalSplit(LiveInterval &VirtReg, AllocationOrder &Order,
   1494                                  SmallVectorImpl<LiveInterval*> &NewVRegs) {
   1495   assert(SA->getUseBlocks().size() == 1 && "Not a local interval");
   1496   const SplitAnalysis::BlockInfo &BI = SA->getUseBlocks().front();
   1497 
   1498   // Note that it is possible to have an interval that is live-in or live-out
   1499   // while only covering a single block - A phi-def can use undef values from
   1500   // predecessors, and the block could be a single-block loop.
   1501   // We don't bother doing anything clever about such a case, we simply assume
   1502   // that the interval is continuous from FirstInstr to LastInstr. We should
   1503   // make sure that we don't do anything illegal to such an interval, though.
   1504 
   1505   ArrayRef<SlotIndex> Uses = SA->getUseSlots();
   1506   if (Uses.size() <= 2)
   1507     return 0;
   1508   const unsigned NumGaps = Uses.size()-1;
   1509 
   1510   DEBUG({
   1511     dbgs() << "tryLocalSplit: ";
   1512     for (unsigned i = 0, e = Uses.size(); i != e; ++i)
   1513       dbgs() << ' ' << Uses[i];
   1514     dbgs() << '\n';
   1515   });
   1516 
   1517   // If VirtReg is live across any register mask operands, compute a list of
   1518   // gaps with register masks.
   1519   SmallVector<unsigned, 8> RegMaskGaps;
   1520   if (Matrix->checkRegMaskInterference(VirtReg)) {
   1521     // Get regmask slots for the whole block.
   1522     ArrayRef<SlotIndex> RMS = LIS->getRegMaskSlotsInBlock(BI.MBB->getNumber());
   1523     DEBUG(dbgs() << RMS.size() << " regmasks in block:");
   1524     // Constrain to VirtReg's live range.
   1525     unsigned ri = std::lower_bound(RMS.begin(), RMS.end(),
   1526                                    Uses.front().getRegSlot()) - RMS.begin();
   1527     unsigned re = RMS.size();
   1528     for (unsigned i = 0; i != NumGaps && ri != re; ++i) {
   1529       // Look for Uses[i] <= RMS <= Uses[i+1].
   1530       assert(!SlotIndex::isEarlierInstr(RMS[ri], Uses[i]));
   1531       if (SlotIndex::isEarlierInstr(Uses[i+1], RMS[ri]))
   1532         continue;
   1533       // Skip a regmask on the same instruction as the last use. It doesn't
   1534       // overlap the live range.
   1535       if (SlotIndex::isSameInstr(Uses[i+1], RMS[ri]) && i+1 == NumGaps)
   1536         break;
   1537       DEBUG(dbgs() << ' ' << RMS[ri] << ':' << Uses[i] << '-' << Uses[i+1]);
   1538       RegMaskGaps.push_back(i);
   1539       // Advance ri to the next gap. A regmask on one of the uses counts in
   1540       // both gaps.
   1541       while (ri != re && SlotIndex::isEarlierInstr(RMS[ri], Uses[i+1]))
   1542         ++ri;
   1543     }
   1544     DEBUG(dbgs() << '\n');
   1545   }
   1546 
   1547   // Since we allow local split results to be split again, there is a risk of
   1548   // creating infinite loops. It is tempting to require that the new live
   1549   // ranges have less instructions than the original. That would guarantee
   1550   // convergence, but it is too strict. A live range with 3 instructions can be
   1551   // split 2+3 (including the COPY), and we want to allow that.
   1552   //
   1553   // Instead we use these rules:
   1554   //
   1555   // 1. Allow any split for ranges with getStage() < RS_Split2. (Except for the
   1556   //    noop split, of course).
   1557   // 2. Require progress be made for ranges with getStage() == RS_Split2. All
   1558   //    the new ranges must have fewer instructions than before the split.
   1559   // 3. New ranges with the same number of instructions are marked RS_Split2,
   1560   //    smaller ranges are marked RS_New.
   1561   //
   1562   // These rules allow a 3 -> 2+3 split once, which we need. They also prevent
   1563   // excessive splitting and infinite loops.
   1564   //
   1565   bool ProgressRequired = getStage(VirtReg) >= RS_Split2;
   1566 
   1567   // Best split candidate.
   1568   unsigned BestBefore = NumGaps;
   1569   unsigned BestAfter = 0;
   1570   float BestDiff = 0;
   1571 
   1572   const float blockFreq =
   1573     SpillPlacer->getBlockFrequency(BI.MBB->getNumber()).getFrequency() *
   1574     (1.0f / BlockFrequency::getEntryFrequency());
   1575   SmallVector<float, 8> GapWeight;
   1576 
   1577   Order.rewind();
   1578   while (unsigned PhysReg = Order.next()) {
   1579     // Keep track of the largest spill weight that would need to be evicted in
   1580     // order to make use of PhysReg between UseSlots[i] and UseSlots[i+1].
   1581     calcGapWeights(PhysReg, GapWeight);
   1582 
   1583     // Remove any gaps with regmask clobbers.
   1584     if (Matrix->checkRegMaskInterference(VirtReg, PhysReg))
   1585       for (unsigned i = 0, e = RegMaskGaps.size(); i != e; ++i)
   1586         GapWeight[RegMaskGaps[i]] = HUGE_VALF;
   1587 
   1588     // Try to find the best sequence of gaps to close.
   1589     // The new spill weight must be larger than any gap interference.
   1590 
   1591     // We will split before Uses[SplitBefore] and after Uses[SplitAfter].
   1592     unsigned SplitBefore = 0, SplitAfter = 1;
   1593 
   1594     // MaxGap should always be max(GapWeight[SplitBefore..SplitAfter-1]).
   1595     // It is the spill weight that needs to be evicted.
   1596     float MaxGap = GapWeight[0];
   1597 
   1598     for (;;) {
   1599       // Live before/after split?
   1600       const bool LiveBefore = SplitBefore != 0 || BI.LiveIn;
   1601       const bool LiveAfter = SplitAfter != NumGaps || BI.LiveOut;
   1602 
   1603       DEBUG(dbgs() << PrintReg(PhysReg, TRI) << ' '
   1604                    << Uses[SplitBefore] << '-' << Uses[SplitAfter]
   1605                    << " i=" << MaxGap);
   1606 
   1607       // Stop before the interval gets so big we wouldn't be making progress.
   1608       if (!LiveBefore && !LiveAfter) {
   1609         DEBUG(dbgs() << " all\n");
   1610         break;
   1611       }
   1612       // Should the interval be extended or shrunk?
   1613       bool Shrink = true;
   1614 
   1615       // How many gaps would the new range have?
   1616       unsigned NewGaps = LiveBefore + SplitAfter - SplitBefore + LiveAfter;
   1617 
   1618       // Legally, without causing looping?
   1619       bool Legal = !ProgressRequired || NewGaps < NumGaps;
   1620 
   1621       if (Legal && MaxGap < HUGE_VALF) {
   1622         // Estimate the new spill weight. Each instruction reads or writes the
   1623         // register. Conservatively assume there are no read-modify-write
   1624         // instructions.
   1625         //
   1626         // Try to guess the size of the new interval.
   1627         const float EstWeight = normalizeSpillWeight(blockFreq * (NewGaps + 1),
   1628                                  Uses[SplitBefore].distance(Uses[SplitAfter]) +
   1629                                  (LiveBefore + LiveAfter)*SlotIndex::InstrDist);
   1630         // Would this split be possible to allocate?
   1631         // Never allocate all gaps, we wouldn't be making progress.
   1632         DEBUG(dbgs() << " w=" << EstWeight);
   1633         if (EstWeight * Hysteresis >= MaxGap) {
   1634           Shrink = false;
   1635           float Diff = EstWeight - MaxGap;
   1636           if (Diff > BestDiff) {
   1637             DEBUG(dbgs() << " (best)");
   1638             BestDiff = Hysteresis * Diff;
   1639             BestBefore = SplitBefore;
   1640             BestAfter = SplitAfter;
   1641           }
   1642         }
   1643       }
   1644 
   1645       // Try to shrink.
   1646       if (Shrink) {
   1647         if (++SplitBefore < SplitAfter) {
   1648           DEBUG(dbgs() << " shrink\n");
   1649           // Recompute the max when necessary.
   1650           if (GapWeight[SplitBefore - 1] >= MaxGap) {
   1651             MaxGap = GapWeight[SplitBefore];
   1652             for (unsigned i = SplitBefore + 1; i != SplitAfter; ++i)
   1653               MaxGap = std::max(MaxGap, GapWeight[i]);
   1654           }
   1655           continue;
   1656         }
   1657         MaxGap = 0;
   1658       }
   1659 
   1660       // Try to extend the interval.
   1661       if (SplitAfter >= NumGaps) {
   1662         DEBUG(dbgs() << " end\n");
   1663         break;
   1664       }
   1665 
   1666       DEBUG(dbgs() << " extend\n");
   1667       MaxGap = std::max(MaxGap, GapWeight[SplitAfter++]);
   1668     }
   1669   }
   1670 
   1671   // Didn't find any candidates?
   1672   if (BestBefore == NumGaps)
   1673     return 0;
   1674 
   1675   DEBUG(dbgs() << "Best local split range: " << Uses[BestBefore]
   1676                << '-' << Uses[BestAfter] << ", " << BestDiff
   1677                << ", " << (BestAfter - BestBefore + 1) << " instrs\n");
   1678 
   1679   LiveRangeEdit LREdit(&VirtReg, NewVRegs, *MF, *LIS, VRM, this);
   1680   SE->reset(LREdit);
   1681 
   1682   SE->openIntv();
   1683   SlotIndex SegStart = SE->enterIntvBefore(Uses[BestBefore]);
   1684   SlotIndex SegStop  = SE->leaveIntvAfter(Uses[BestAfter]);
   1685   SE->useIntv(SegStart, SegStop);
   1686   SmallVector<unsigned, 8> IntvMap;
   1687   SE->finish(&IntvMap);
   1688   DebugVars->splitRegister(VirtReg.reg, LREdit.regs());
   1689 
   1690   // If the new range has the same number of instructions as before, mark it as
   1691   // RS_Split2 so the next split will be forced to make progress. Otherwise,
   1692   // leave the new intervals as RS_New so they can compete.
   1693   bool LiveBefore = BestBefore != 0 || BI.LiveIn;
   1694   bool LiveAfter = BestAfter != NumGaps || BI.LiveOut;
   1695   unsigned NewGaps = LiveBefore + BestAfter - BestBefore + LiveAfter;
   1696   if (NewGaps >= NumGaps) {
   1697     DEBUG(dbgs() << "Tagging non-progress ranges: ");
   1698     assert(!ProgressRequired && "Didn't make progress when it was required.");
   1699     for (unsigned i = 0, e = IntvMap.size(); i != e; ++i)
   1700       if (IntvMap[i] == 1) {
   1701         setStage(*LREdit.get(i), RS_Split2);
   1702         DEBUG(dbgs() << PrintReg(LREdit.get(i)->reg));
   1703       }
   1704     DEBUG(dbgs() << '\n');
   1705   }
   1706   ++NumLocalSplits;
   1707 
   1708   return 0;
   1709 }
   1710 
   1711 //===----------------------------------------------------------------------===//
   1712 //                          Live Range Splitting
   1713 //===----------------------------------------------------------------------===//
   1714 
   1715 /// trySplit - Try to split VirtReg or one of its interferences, making it
   1716 /// assignable.
   1717 /// @return Physreg when VirtReg may be assigned and/or new NewVRegs.
   1718 unsigned RAGreedy::trySplit(LiveInterval &VirtReg, AllocationOrder &Order,
   1719                             SmallVectorImpl<LiveInterval*>&NewVRegs) {
   1720   // Ranges must be Split2 or less.
   1721   if (getStage(VirtReg) >= RS_Spill)
   1722     return 0;
   1723 
   1724   // Local intervals are handled separately.
   1725   if (LIS->intervalIsInOneMBB(VirtReg)) {
   1726     NamedRegionTimer T("Local Splitting", TimerGroupName, TimePassesIsEnabled);
   1727     SA->analyze(&VirtReg);
   1728     unsigned PhysReg = tryLocalSplit(VirtReg, Order, NewVRegs);
   1729     if (PhysReg || !NewVRegs.empty())
   1730       return PhysReg;
   1731     return tryInstructionSplit(VirtReg, Order, NewVRegs);
   1732   }
   1733 
   1734   NamedRegionTimer T("Global Splitting", TimerGroupName, TimePassesIsEnabled);
   1735 
   1736   SA->analyze(&VirtReg);
   1737 
   1738   // FIXME: SplitAnalysis may repair broken live ranges coming from the
   1739   // coalescer. That may cause the range to become allocatable which means that
   1740   // tryRegionSplit won't be making progress. This check should be replaced with
   1741   // an assertion when the coalescer is fixed.
   1742   if (SA->didRepairRange()) {
   1743     // VirtReg has changed, so all cached queries are invalid.
   1744     Matrix->invalidateVirtRegs();
   1745     if (unsigned PhysReg = tryAssign(VirtReg, Order, NewVRegs))
   1746       return PhysReg;
   1747   }
   1748 
   1749   // First try to split around a region spanning multiple blocks. RS_Split2
   1750   // ranges already made dubious progress with region splitting, so they go
   1751   // straight to single block splitting.
   1752   if (getStage(VirtReg) < RS_Split2) {
   1753     unsigned PhysReg = tryRegionSplit(VirtReg, Order, NewVRegs);
   1754     if (PhysReg || !NewVRegs.empty())
   1755       return PhysReg;
   1756   }
   1757 
   1758   // Then isolate blocks.
   1759   return tryBlockSplit(VirtReg, Order, NewVRegs);
   1760 }
   1761 
   1762 
   1763 //===----------------------------------------------------------------------===//
   1764 //                            Main Entry Point
   1765 //===----------------------------------------------------------------------===//
   1766 
   1767 unsigned RAGreedy::selectOrSplit(LiveInterval &VirtReg,
   1768                                  SmallVectorImpl<LiveInterval*> &NewVRegs) {
   1769   // First try assigning a free register.
   1770   AllocationOrder Order(VirtReg.reg, *VRM, RegClassInfo);
   1771   if (unsigned PhysReg = tryAssign(VirtReg, Order, NewVRegs))
   1772     return PhysReg;
   1773 
   1774   LiveRangeStage Stage = getStage(VirtReg);
   1775   DEBUG(dbgs() << StageName[Stage]
   1776                << " Cascade " << ExtraRegInfo[VirtReg.reg].Cascade << '\n');
   1777 
   1778   // Try to evict a less worthy live range, but only for ranges from the primary
   1779   // queue. The RS_Split ranges already failed to do this, and they should not
   1780   // get a second chance until they have been split.
   1781   if (Stage != RS_Split)
   1782     if (unsigned PhysReg = tryEvict(VirtReg, Order, NewVRegs))
   1783       return PhysReg;
   1784 
   1785   assert(NewVRegs.empty() && "Cannot append to existing NewVRegs");
   1786 
   1787   // The first time we see a live range, don't try to split or spill.
   1788   // Wait until the second time, when all smaller ranges have been allocated.
   1789   // This gives a better picture of the interference to split around.
   1790   if (Stage < RS_Split) {
   1791     setStage(VirtReg, RS_Split);
   1792     DEBUG(dbgs() << "wait for second round\n");
   1793     NewVRegs.push_back(&VirtReg);
   1794     return 0;
   1795   }
   1796 
   1797   // If we couldn't allocate a register from spilling, there is probably some
   1798   // invalid inline assembly. The base class wil report it.
   1799   if (Stage >= RS_Done || !VirtReg.isSpillable())
   1800     return ~0u;
   1801 
   1802   // Try splitting VirtReg or interferences.
   1803   unsigned PhysReg = trySplit(VirtReg, Order, NewVRegs);
   1804   if (PhysReg || !NewVRegs.empty())
   1805     return PhysReg;
   1806 
   1807   // Finally spill VirtReg itself.
   1808   NamedRegionTimer T("Spiller", TimerGroupName, TimePassesIsEnabled);
   1809   LiveRangeEdit LRE(&VirtReg, NewVRegs, *MF, *LIS, VRM, this);
   1810   spiller().spill(LRE);
   1811   setStage(NewVRegs.begin(), NewVRegs.end(), RS_Done);
   1812 
   1813   if (VerifyEnabled)
   1814     MF->verify(this, "After spilling");
   1815 
   1816   // The live virtual register requesting allocation was spilled, so tell
   1817   // the caller not to allocate anything during this round.
   1818   return 0;
   1819 }
   1820 
   1821 bool RAGreedy::runOnMachineFunction(MachineFunction &mf) {
   1822   DEBUG(dbgs() << "********** GREEDY REGISTER ALLOCATION **********\n"
   1823                << "********** Function: " << mf.getName() << '\n');
   1824 
   1825   MF = &mf;
   1826   if (VerifyEnabled)
   1827     MF->verify(this, "Before greedy register allocator");
   1828 
   1829   RegAllocBase::init(getAnalysis<VirtRegMap>(),
   1830                      getAnalysis<LiveIntervals>(),
   1831                      getAnalysis<LiveRegMatrix>());
   1832   Indexes = &getAnalysis<SlotIndexes>();
   1833   MBFI = &getAnalysis<MachineBlockFrequencyInfo>();
   1834   DomTree = &getAnalysis<MachineDominatorTree>();
   1835   SpillerInstance.reset(createInlineSpiller(*this, *MF, *VRM));
   1836   Loops = &getAnalysis<MachineLoopInfo>();
   1837   Bundles = &getAnalysis<EdgeBundles>();
   1838   SpillPlacer = &getAnalysis<SpillPlacement>();
   1839   DebugVars = &getAnalysis<LiveDebugVariables>();
   1840 
   1841   DEBUG(LIS->dump());
   1842 
   1843   SA.reset(new SplitAnalysis(*VRM, *LIS, *Loops));
   1844   SE.reset(new SplitEditor(*SA, *LIS, *VRM, *DomTree, *MBFI));
   1845   ExtraRegInfo.clear();
   1846   ExtraRegInfo.resize(MRI->getNumVirtRegs());
   1847   NextCascade = 1;
   1848   IntfCache.init(MF, Matrix->getLiveUnions(), Indexes, LIS, TRI);
   1849   GlobalCand.resize(32);  // This will grow as needed.
   1850 
   1851   allocatePhysRegs();
   1852   releaseMemory();
   1853   return true;
   1854 }
   1855