ScheduleDAGInstrs.cpp revision 353358
1//===---- ScheduleDAGInstrs.cpp - MachineInstr Rescheduling ---------------===//
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/// \file This implements the ScheduleDAGInstrs class, which implements
10/// re-scheduling of MachineInstrs.
11//
12//===----------------------------------------------------------------------===//
13
14#include "llvm/CodeGen/ScheduleDAGInstrs.h"
15#include "llvm/ADT/IntEqClasses.h"
16#include "llvm/ADT/MapVector.h"
17#include "llvm/ADT/SmallPtrSet.h"
18#include "llvm/ADT/SmallVector.h"
19#include "llvm/ADT/SparseSet.h"
20#include "llvm/ADT/iterator_range.h"
21#include "llvm/Analysis/AliasAnalysis.h"
22#include "llvm/Analysis/ValueTracking.h"
23#include "llvm/CodeGen/LiveIntervals.h"
24#include "llvm/CodeGen/LivePhysRegs.h"
25#include "llvm/CodeGen/MachineBasicBlock.h"
26#include "llvm/CodeGen/MachineFrameInfo.h"
27#include "llvm/CodeGen/MachineFunction.h"
28#include "llvm/CodeGen/MachineInstr.h"
29#include "llvm/CodeGen/MachineInstrBundle.h"
30#include "llvm/CodeGen/MachineMemOperand.h"
31#include "llvm/CodeGen/MachineOperand.h"
32#include "llvm/CodeGen/MachineRegisterInfo.h"
33#include "llvm/CodeGen/PseudoSourceValue.h"
34#include "llvm/CodeGen/RegisterPressure.h"
35#include "llvm/CodeGen/ScheduleDAG.h"
36#include "llvm/CodeGen/ScheduleDFS.h"
37#include "llvm/CodeGen/SlotIndexes.h"
38#include "llvm/CodeGen/TargetRegisterInfo.h"
39#include "llvm/CodeGen/TargetSubtargetInfo.h"
40#include "llvm/Config/llvm-config.h"
41#include "llvm/IR/Constants.h"
42#include "llvm/IR/Function.h"
43#include "llvm/IR/Instruction.h"
44#include "llvm/IR/Instructions.h"
45#include "llvm/IR/Operator.h"
46#include "llvm/IR/Type.h"
47#include "llvm/IR/Value.h"
48#include "llvm/MC/LaneBitmask.h"
49#include "llvm/MC/MCRegisterInfo.h"
50#include "llvm/Support/Casting.h"
51#include "llvm/Support/CommandLine.h"
52#include "llvm/Support/Compiler.h"
53#include "llvm/Support/Debug.h"
54#include "llvm/Support/ErrorHandling.h"
55#include "llvm/Support/Format.h"
56#include "llvm/Support/raw_ostream.h"
57#include <algorithm>
58#include <cassert>
59#include <iterator>
60#include <string>
61#include <utility>
62#include <vector>
63
64using namespace llvm;
65
66#define DEBUG_TYPE "machine-scheduler"
67
68static cl::opt<bool> EnableAASchedMI("enable-aa-sched-mi", cl::Hidden,
69    cl::ZeroOrMore, cl::init(false),
70    cl::desc("Enable use of AA during MI DAG construction"));
71
72static cl::opt<bool> UseTBAA("use-tbaa-in-sched-mi", cl::Hidden,
73    cl::init(true), cl::desc("Enable use of TBAA during MI DAG construction"));
74
75// Note: the two options below might be used in tuning compile time vs
76// output quality. Setting HugeRegion so large that it will never be
77// reached means best-effort, but may be slow.
78
79// When Stores and Loads maps (or NonAliasStores and NonAliasLoads)
80// together hold this many SUs, a reduction of maps will be done.
81static cl::opt<unsigned> HugeRegion("dag-maps-huge-region", cl::Hidden,
82    cl::init(1000), cl::desc("The limit to use while constructing the DAG "
83                             "prior to scheduling, at which point a trade-off "
84                             "is made to avoid excessive compile time."));
85
86static cl::opt<unsigned> ReductionSize(
87    "dag-maps-reduction-size", cl::Hidden,
88    cl::desc("A huge scheduling region will have maps reduced by this many "
89             "nodes at a time. Defaults to HugeRegion / 2."));
90
91static unsigned getReductionSize() {
92  // Always reduce a huge region with half of the elements, except
93  // when user sets this number explicitly.
94  if (ReductionSize.getNumOccurrences() == 0)
95    return HugeRegion / 2;
96  return ReductionSize;
97}
98
99static void dumpSUList(ScheduleDAGInstrs::SUList &L) {
100#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
101  dbgs() << "{ ";
102  for (const SUnit *su : L) {
103    dbgs() << "SU(" << su->NodeNum << ")";
104    if (su != L.back())
105      dbgs() << ", ";
106  }
107  dbgs() << "}\n";
108#endif
109}
110
111ScheduleDAGInstrs::ScheduleDAGInstrs(MachineFunction &mf,
112                                     const MachineLoopInfo *mli,
113                                     bool RemoveKillFlags)
114    : ScheduleDAG(mf), MLI(mli), MFI(mf.getFrameInfo()),
115      RemoveKillFlags(RemoveKillFlags),
116      UnknownValue(UndefValue::get(
117                             Type::getVoidTy(mf.getFunction().getContext()))), Topo(SUnits, &ExitSU) {
118  DbgValues.clear();
119
120  const TargetSubtargetInfo &ST = mf.getSubtarget();
121  SchedModel.init(&ST);
122}
123
124/// If this machine instr has memory reference information and it can be
125/// tracked to a normal reference to a known object, return the Value
126/// for that object. This function returns false the memory location is
127/// unknown or may alias anything.
128static bool getUnderlyingObjectsForInstr(const MachineInstr *MI,
129                                         const MachineFrameInfo &MFI,
130                                         UnderlyingObjectsVector &Objects,
131                                         const DataLayout &DL) {
132  auto allMMOsOkay = [&]() {
133    for (const MachineMemOperand *MMO : MI->memoperands()) {
134      // TODO: Figure out whether isAtomic is really necessary (see D57601).
135      if (MMO->isVolatile() || MMO->isAtomic())
136        return false;
137
138      if (const PseudoSourceValue *PSV = MMO->getPseudoValue()) {
139        // Function that contain tail calls don't have unique PseudoSourceValue
140        // objects. Two PseudoSourceValues might refer to the same or
141        // overlapping locations. The client code calling this function assumes
142        // this is not the case. So return a conservative answer of no known
143        // object.
144        if (MFI.hasTailCall())
145          return false;
146
147        // For now, ignore PseudoSourceValues which may alias LLVM IR values
148        // because the code that uses this function has no way to cope with
149        // such aliases.
150        if (PSV->isAliased(&MFI))
151          return false;
152
153        bool MayAlias = PSV->mayAlias(&MFI);
154        Objects.push_back(UnderlyingObjectsVector::value_type(PSV, MayAlias));
155      } else if (const Value *V = MMO->getValue()) {
156        SmallVector<Value *, 4> Objs;
157        if (!getUnderlyingObjectsForCodeGen(V, Objs, DL))
158          return false;
159
160        for (Value *V : Objs) {
161          assert(isIdentifiedObject(V));
162          Objects.push_back(UnderlyingObjectsVector::value_type(V, true));
163        }
164      } else
165        return false;
166    }
167    return true;
168  };
169
170  if (!allMMOsOkay()) {
171    Objects.clear();
172    return false;
173  }
174
175  return true;
176}
177
178void ScheduleDAGInstrs::startBlock(MachineBasicBlock *bb) {
179  BB = bb;
180}
181
182void ScheduleDAGInstrs::finishBlock() {
183  // Subclasses should no longer refer to the old block.
184  BB = nullptr;
185}
186
187void ScheduleDAGInstrs::enterRegion(MachineBasicBlock *bb,
188                                    MachineBasicBlock::iterator begin,
189                                    MachineBasicBlock::iterator end,
190                                    unsigned regioninstrs) {
191  assert(bb == BB && "startBlock should set BB");
192  RegionBegin = begin;
193  RegionEnd = end;
194  NumRegionInstrs = regioninstrs;
195}
196
197void ScheduleDAGInstrs::exitRegion() {
198  // Nothing to do.
199}
200
201void ScheduleDAGInstrs::addSchedBarrierDeps() {
202  MachineInstr *ExitMI = RegionEnd != BB->end() ? &*RegionEnd : nullptr;
203  ExitSU.setInstr(ExitMI);
204  // Add dependencies on the defs and uses of the instruction.
205  if (ExitMI) {
206    for (const MachineOperand &MO : ExitMI->operands()) {
207      if (!MO.isReg() || MO.isDef()) continue;
208      unsigned Reg = MO.getReg();
209      if (TargetRegisterInfo::isPhysicalRegister(Reg)) {
210        Uses.insert(PhysRegSUOper(&ExitSU, -1, Reg));
211      } else if (TargetRegisterInfo::isVirtualRegister(Reg) && MO.readsReg()) {
212        addVRegUseDeps(&ExitSU, ExitMI->getOperandNo(&MO));
213      }
214    }
215  }
216  if (!ExitMI || (!ExitMI->isCall() && !ExitMI->isBarrier())) {
217    // For others, e.g. fallthrough, conditional branch, assume the exit
218    // uses all the registers that are livein to the successor blocks.
219    for (const MachineBasicBlock *Succ : BB->successors()) {
220      for (const auto &LI : Succ->liveins()) {
221        if (!Uses.contains(LI.PhysReg))
222          Uses.insert(PhysRegSUOper(&ExitSU, -1, LI.PhysReg));
223      }
224    }
225  }
226}
227
228/// MO is an operand of SU's instruction that defines a physical register. Adds
229/// data dependencies from SU to any uses of the physical register.
230void ScheduleDAGInstrs::addPhysRegDataDeps(SUnit *SU, unsigned OperIdx) {
231  const MachineOperand &MO = SU->getInstr()->getOperand(OperIdx);
232  assert(MO.isDef() && "expect physreg def");
233
234  // Ask the target if address-backscheduling is desirable, and if so how much.
235  const TargetSubtargetInfo &ST = MF.getSubtarget();
236
237  // Only use any non-zero latency for real defs/uses, in contrast to
238  // "fake" operands added by regalloc.
239  const MCInstrDesc *DefMIDesc = &SU->getInstr()->getDesc();
240  bool ImplicitPseudoDef = (OperIdx >= DefMIDesc->getNumOperands() &&
241                            !DefMIDesc->hasImplicitDefOfPhysReg(MO.getReg()));
242  for (MCRegAliasIterator Alias(MO.getReg(), TRI, true);
243       Alias.isValid(); ++Alias) {
244    if (!Uses.contains(*Alias))
245      continue;
246    for (Reg2SUnitsMap::iterator I = Uses.find(*Alias); I != Uses.end(); ++I) {
247      SUnit *UseSU = I->SU;
248      if (UseSU == SU)
249        continue;
250
251      // Adjust the dependence latency using operand def/use information,
252      // then allow the target to perform its own adjustments.
253      int UseOp = I->OpIdx;
254      MachineInstr *RegUse = nullptr;
255      SDep Dep;
256      if (UseOp < 0)
257        Dep = SDep(SU, SDep::Artificial);
258      else {
259        // Set the hasPhysRegDefs only for physreg defs that have a use within
260        // the scheduling region.
261        SU->hasPhysRegDefs = true;
262        Dep = SDep(SU, SDep::Data, *Alias);
263        RegUse = UseSU->getInstr();
264      }
265      const MCInstrDesc *UseMIDesc =
266          (RegUse ? &UseSU->getInstr()->getDesc() : nullptr);
267      bool ImplicitPseudoUse =
268          (UseMIDesc && UseOp >= ((int)UseMIDesc->getNumOperands()) &&
269           !UseMIDesc->hasImplicitUseOfPhysReg(*Alias));
270      if (!ImplicitPseudoDef && !ImplicitPseudoUse) {
271        Dep.setLatency(SchedModel.computeOperandLatency(SU->getInstr(), OperIdx,
272                                                        RegUse, UseOp));
273        ST.adjustSchedDependency(SU, UseSU, Dep);
274      } else
275        Dep.setLatency(0);
276
277      UseSU->addPred(Dep);
278    }
279  }
280}
281
282/// Adds register dependencies (data, anti, and output) from this SUnit
283/// to following instructions in the same scheduling region that depend the
284/// physical register referenced at OperIdx.
285void ScheduleDAGInstrs::addPhysRegDeps(SUnit *SU, unsigned OperIdx) {
286  MachineInstr *MI = SU->getInstr();
287  MachineOperand &MO = MI->getOperand(OperIdx);
288  unsigned Reg = MO.getReg();
289  // We do not need to track any dependencies for constant registers.
290  if (MRI.isConstantPhysReg(Reg))
291    return;
292
293  // Optionally add output and anti dependencies. For anti
294  // dependencies we use a latency of 0 because for a multi-issue
295  // target we want to allow the defining instruction to issue
296  // in the same cycle as the using instruction.
297  // TODO: Using a latency of 1 here for output dependencies assumes
298  //       there's no cost for reusing registers.
299  SDep::Kind Kind = MO.isUse() ? SDep::Anti : SDep::Output;
300  for (MCRegAliasIterator Alias(Reg, TRI, true); Alias.isValid(); ++Alias) {
301    if (!Defs.contains(*Alias))
302      continue;
303    for (Reg2SUnitsMap::iterator I = Defs.find(*Alias); I != Defs.end(); ++I) {
304      SUnit *DefSU = I->SU;
305      if (DefSU == &ExitSU)
306        continue;
307      if (DefSU != SU &&
308          (Kind != SDep::Output || !MO.isDead() ||
309           !DefSU->getInstr()->registerDefIsDead(*Alias))) {
310        if (Kind == SDep::Anti)
311          DefSU->addPred(SDep(SU, Kind, /*Reg=*/*Alias));
312        else {
313          SDep Dep(SU, Kind, /*Reg=*/*Alias);
314          Dep.setLatency(
315            SchedModel.computeOutputLatency(MI, OperIdx, DefSU->getInstr()));
316          DefSU->addPred(Dep);
317        }
318      }
319    }
320  }
321
322  if (!MO.isDef()) {
323    SU->hasPhysRegUses = true;
324    // Either insert a new Reg2SUnits entry with an empty SUnits list, or
325    // retrieve the existing SUnits list for this register's uses.
326    // Push this SUnit on the use list.
327    Uses.insert(PhysRegSUOper(SU, OperIdx, Reg));
328    if (RemoveKillFlags)
329      MO.setIsKill(false);
330  } else {
331    addPhysRegDataDeps(SU, OperIdx);
332
333    // Clear previous uses and defs of this register and its subergisters.
334    for (MCSubRegIterator SubReg(Reg, TRI, true); SubReg.isValid(); ++SubReg) {
335      if (Uses.contains(*SubReg))
336        Uses.eraseAll(*SubReg);
337      if (!MO.isDead())
338        Defs.eraseAll(*SubReg);
339    }
340    if (MO.isDead() && SU->isCall) {
341      // Calls will not be reordered because of chain dependencies (see
342      // below). Since call operands are dead, calls may continue to be added
343      // to the DefList making dependence checking quadratic in the size of
344      // the block. Instead, we leave only one call at the back of the
345      // DefList.
346      Reg2SUnitsMap::RangePair P = Defs.equal_range(Reg);
347      Reg2SUnitsMap::iterator B = P.first;
348      Reg2SUnitsMap::iterator I = P.second;
349      for (bool isBegin = I == B; !isBegin; /* empty */) {
350        isBegin = (--I) == B;
351        if (!I->SU->isCall)
352          break;
353        I = Defs.erase(I);
354      }
355    }
356
357    // Defs are pushed in the order they are visited and never reordered.
358    Defs.insert(PhysRegSUOper(SU, OperIdx, Reg));
359  }
360}
361
362LaneBitmask ScheduleDAGInstrs::getLaneMaskForMO(const MachineOperand &MO) const
363{
364  unsigned Reg = MO.getReg();
365  // No point in tracking lanemasks if we don't have interesting subregisters.
366  const TargetRegisterClass &RC = *MRI.getRegClass(Reg);
367  if (!RC.HasDisjunctSubRegs)
368    return LaneBitmask::getAll();
369
370  unsigned SubReg = MO.getSubReg();
371  if (SubReg == 0)
372    return RC.getLaneMask();
373  return TRI->getSubRegIndexLaneMask(SubReg);
374}
375
376/// Adds register output and data dependencies from this SUnit to instructions
377/// that occur later in the same scheduling region if they read from or write to
378/// the virtual register defined at OperIdx.
379///
380/// TODO: Hoist loop induction variable increments. This has to be
381/// reevaluated. Generally, IV scheduling should be done before coalescing.
382void ScheduleDAGInstrs::addVRegDefDeps(SUnit *SU, unsigned OperIdx) {
383  MachineInstr *MI = SU->getInstr();
384  MachineOperand &MO = MI->getOperand(OperIdx);
385  unsigned Reg = MO.getReg();
386
387  LaneBitmask DefLaneMask;
388  LaneBitmask KillLaneMask;
389  if (TrackLaneMasks) {
390    bool IsKill = MO.getSubReg() == 0 || MO.isUndef();
391    DefLaneMask = getLaneMaskForMO(MO);
392    // If we have a <read-undef> flag, none of the lane values comes from an
393    // earlier instruction.
394    KillLaneMask = IsKill ? LaneBitmask::getAll() : DefLaneMask;
395
396    // Clear undef flag, we'll re-add it later once we know which subregister
397    // Def is first.
398    MO.setIsUndef(false);
399  } else {
400    DefLaneMask = LaneBitmask::getAll();
401    KillLaneMask = LaneBitmask::getAll();
402  }
403
404  if (MO.isDead()) {
405    assert(CurrentVRegUses.find(Reg) == CurrentVRegUses.end() &&
406           "Dead defs should have no uses");
407  } else {
408    // Add data dependence to all uses we found so far.
409    const TargetSubtargetInfo &ST = MF.getSubtarget();
410    for (VReg2SUnitOperIdxMultiMap::iterator I = CurrentVRegUses.find(Reg),
411         E = CurrentVRegUses.end(); I != E; /*empty*/) {
412      LaneBitmask LaneMask = I->LaneMask;
413      // Ignore uses of other lanes.
414      if ((LaneMask & KillLaneMask).none()) {
415        ++I;
416        continue;
417      }
418
419      if ((LaneMask & DefLaneMask).any()) {
420        SUnit *UseSU = I->SU;
421        MachineInstr *Use = UseSU->getInstr();
422        SDep Dep(SU, SDep::Data, Reg);
423        Dep.setLatency(SchedModel.computeOperandLatency(MI, OperIdx, Use,
424                                                        I->OperandIndex));
425        ST.adjustSchedDependency(SU, UseSU, Dep);
426        UseSU->addPred(Dep);
427      }
428
429      LaneMask &= ~KillLaneMask;
430      // If we found a Def for all lanes of this use, remove it from the list.
431      if (LaneMask.any()) {
432        I->LaneMask = LaneMask;
433        ++I;
434      } else
435        I = CurrentVRegUses.erase(I);
436    }
437  }
438
439  // Shortcut: Singly defined vregs do not have output/anti dependencies.
440  if (MRI.hasOneDef(Reg))
441    return;
442
443  // Add output dependence to the next nearest defs of this vreg.
444  //
445  // Unless this definition is dead, the output dependence should be
446  // transitively redundant with antidependencies from this definition's
447  // uses. We're conservative for now until we have a way to guarantee the uses
448  // are not eliminated sometime during scheduling. The output dependence edge
449  // is also useful if output latency exceeds def-use latency.
450  LaneBitmask LaneMask = DefLaneMask;
451  for (VReg2SUnit &V2SU : make_range(CurrentVRegDefs.find(Reg),
452                                     CurrentVRegDefs.end())) {
453    // Ignore defs for other lanes.
454    if ((V2SU.LaneMask & LaneMask).none())
455      continue;
456    // Add an output dependence.
457    SUnit *DefSU = V2SU.SU;
458    // Ignore additional defs of the same lanes in one instruction. This can
459    // happen because lanemasks are shared for targets with too many
460    // subregisters. We also use some representration tricks/hacks where we
461    // add super-register defs/uses, to imply that although we only access parts
462    // of the reg we care about the full one.
463    if (DefSU == SU)
464      continue;
465    SDep Dep(SU, SDep::Output, Reg);
466    Dep.setLatency(
467      SchedModel.computeOutputLatency(MI, OperIdx, DefSU->getInstr()));
468    DefSU->addPred(Dep);
469
470    // Update current definition. This can get tricky if the def was about a
471    // bigger lanemask before. We then have to shrink it and create a new
472    // VReg2SUnit for the non-overlapping part.
473    LaneBitmask OverlapMask = V2SU.LaneMask & LaneMask;
474    LaneBitmask NonOverlapMask = V2SU.LaneMask & ~LaneMask;
475    V2SU.SU = SU;
476    V2SU.LaneMask = OverlapMask;
477    if (NonOverlapMask.any())
478      CurrentVRegDefs.insert(VReg2SUnit(Reg, NonOverlapMask, DefSU));
479  }
480  // If there was no CurrentVRegDefs entry for some lanes yet, create one.
481  if (LaneMask.any())
482    CurrentVRegDefs.insert(VReg2SUnit(Reg, LaneMask, SU));
483}
484
485/// Adds a register data dependency if the instruction that defines the
486/// virtual register used at OperIdx is mapped to an SUnit. Add a register
487/// antidependency from this SUnit to instructions that occur later in the same
488/// scheduling region if they write the virtual register.
489///
490/// TODO: Handle ExitSU "uses" properly.
491void ScheduleDAGInstrs::addVRegUseDeps(SUnit *SU, unsigned OperIdx) {
492  const MachineInstr *MI = SU->getInstr();
493  const MachineOperand &MO = MI->getOperand(OperIdx);
494  unsigned Reg = MO.getReg();
495
496  // Remember the use. Data dependencies will be added when we find the def.
497  LaneBitmask LaneMask = TrackLaneMasks ? getLaneMaskForMO(MO)
498                                        : LaneBitmask::getAll();
499  CurrentVRegUses.insert(VReg2SUnitOperIdx(Reg, LaneMask, OperIdx, SU));
500
501  // Add antidependences to the following defs of the vreg.
502  for (VReg2SUnit &V2SU : make_range(CurrentVRegDefs.find(Reg),
503                                     CurrentVRegDefs.end())) {
504    // Ignore defs for unrelated lanes.
505    LaneBitmask PrevDefLaneMask = V2SU.LaneMask;
506    if ((PrevDefLaneMask & LaneMask).none())
507      continue;
508    if (V2SU.SU == SU)
509      continue;
510
511    V2SU.SU->addPred(SDep(SU, SDep::Anti, Reg));
512  }
513}
514
515/// Returns true if MI is an instruction we are unable to reason about
516/// (like a call or something with unmodeled side effects).
517static inline bool isGlobalMemoryObject(AliasAnalysis *AA, MachineInstr *MI) {
518  return MI->isCall() || MI->hasUnmodeledSideEffects() ||
519         (MI->hasOrderedMemoryRef() && !MI->isDereferenceableInvariantLoad(AA));
520}
521
522void ScheduleDAGInstrs::addChainDependency (SUnit *SUa, SUnit *SUb,
523                                            unsigned Latency) {
524  if (SUa->getInstr()->mayAlias(AAForDep, *SUb->getInstr(), UseTBAA)) {
525    SDep Dep(SUa, SDep::MayAliasMem);
526    Dep.setLatency(Latency);
527    SUb->addPred(Dep);
528  }
529}
530
531/// Creates an SUnit for each real instruction, numbered in top-down
532/// topological order. The instruction order A < B, implies that no edge exists
533/// from B to A.
534///
535/// Map each real instruction to its SUnit.
536///
537/// After initSUnits, the SUnits vector cannot be resized and the scheduler may
538/// hang onto SUnit pointers. We may relax this in the future by using SUnit IDs
539/// instead of pointers.
540///
541/// MachineScheduler relies on initSUnits numbering the nodes by their order in
542/// the original instruction list.
543void ScheduleDAGInstrs::initSUnits() {
544  // We'll be allocating one SUnit for each real instruction in the region,
545  // which is contained within a basic block.
546  SUnits.reserve(NumRegionInstrs);
547
548  for (MachineInstr &MI : make_range(RegionBegin, RegionEnd)) {
549    if (MI.isDebugInstr())
550      continue;
551
552    SUnit *SU = newSUnit(&MI);
553    MISUnitMap[&MI] = SU;
554
555    SU->isCall = MI.isCall();
556    SU->isCommutable = MI.isCommutable();
557
558    // Assign the Latency field of SU using target-provided information.
559    SU->Latency = SchedModel.computeInstrLatency(SU->getInstr());
560
561    // If this SUnit uses a reserved or unbuffered resource, mark it as such.
562    //
563    // Reserved resources block an instruction from issuing and stall the
564    // entire pipeline. These are identified by BufferSize=0.
565    //
566    // Unbuffered resources prevent execution of subsequent instructions that
567    // require the same resources. This is used for in-order execution pipelines
568    // within an out-of-order core. These are identified by BufferSize=1.
569    if (SchedModel.hasInstrSchedModel()) {
570      const MCSchedClassDesc *SC = getSchedClass(SU);
571      for (const MCWriteProcResEntry &PRE :
572           make_range(SchedModel.getWriteProcResBegin(SC),
573                      SchedModel.getWriteProcResEnd(SC))) {
574        switch (SchedModel.getProcResource(PRE.ProcResourceIdx)->BufferSize) {
575        case 0:
576          SU->hasReservedResource = true;
577          break;
578        case 1:
579          SU->isUnbuffered = true;
580          break;
581        default:
582          break;
583        }
584      }
585    }
586  }
587}
588
589class ScheduleDAGInstrs::Value2SUsMap : public MapVector<ValueType, SUList> {
590  /// Current total number of SUs in map.
591  unsigned NumNodes = 0;
592
593  /// 1 for loads, 0 for stores. (see comment in SUList)
594  unsigned TrueMemOrderLatency;
595
596public:
597  Value2SUsMap(unsigned lat = 0) : TrueMemOrderLatency(lat) {}
598
599  /// To keep NumNodes up to date, insert() is used instead of
600  /// this operator w/ push_back().
601  ValueType &operator[](const SUList &Key) {
602    llvm_unreachable("Don't use. Use insert() instead."); };
603
604  /// Adds SU to the SUList of V. If Map grows huge, reduce its size by calling
605  /// reduce().
606  void inline insert(SUnit *SU, ValueType V) {
607    MapVector::operator[](V).push_back(SU);
608    NumNodes++;
609  }
610
611  /// Clears the list of SUs mapped to V.
612  void inline clearList(ValueType V) {
613    iterator Itr = find(V);
614    if (Itr != end()) {
615      assert(NumNodes >= Itr->second.size());
616      NumNodes -= Itr->second.size();
617
618      Itr->second.clear();
619    }
620  }
621
622  /// Clears map from all contents.
623  void clear() {
624    MapVector<ValueType, SUList>::clear();
625    NumNodes = 0;
626  }
627
628  unsigned inline size() const { return NumNodes; }
629
630  /// Counts the number of SUs in this map after a reduction.
631  void reComputeSize() {
632    NumNodes = 0;
633    for (auto &I : *this)
634      NumNodes += I.second.size();
635  }
636
637  unsigned inline getTrueMemOrderLatency() const {
638    return TrueMemOrderLatency;
639  }
640
641  void dump();
642};
643
644void ScheduleDAGInstrs::addChainDependencies(SUnit *SU,
645                                             Value2SUsMap &Val2SUsMap) {
646  for (auto &I : Val2SUsMap)
647    addChainDependencies(SU, I.second,
648                         Val2SUsMap.getTrueMemOrderLatency());
649}
650
651void ScheduleDAGInstrs::addChainDependencies(SUnit *SU,
652                                             Value2SUsMap &Val2SUsMap,
653                                             ValueType V) {
654  Value2SUsMap::iterator Itr = Val2SUsMap.find(V);
655  if (Itr != Val2SUsMap.end())
656    addChainDependencies(SU, Itr->second,
657                         Val2SUsMap.getTrueMemOrderLatency());
658}
659
660void ScheduleDAGInstrs::addBarrierChain(Value2SUsMap &map) {
661  assert(BarrierChain != nullptr);
662
663  for (auto &I : map) {
664    SUList &sus = I.second;
665    for (auto *SU : sus)
666      SU->addPredBarrier(BarrierChain);
667  }
668  map.clear();
669}
670
671void ScheduleDAGInstrs::insertBarrierChain(Value2SUsMap &map) {
672  assert(BarrierChain != nullptr);
673
674  // Go through all lists of SUs.
675  for (Value2SUsMap::iterator I = map.begin(), EE = map.end(); I != EE;) {
676    Value2SUsMap::iterator CurrItr = I++;
677    SUList &sus = CurrItr->second;
678    SUList::iterator SUItr = sus.begin(), SUEE = sus.end();
679    for (; SUItr != SUEE; ++SUItr) {
680      // Stop on BarrierChain or any instruction above it.
681      if ((*SUItr)->NodeNum <= BarrierChain->NodeNum)
682        break;
683
684      (*SUItr)->addPredBarrier(BarrierChain);
685    }
686
687    // Remove also the BarrierChain from list if present.
688    if (SUItr != SUEE && *SUItr == BarrierChain)
689      SUItr++;
690
691    // Remove all SUs that are now successors of BarrierChain.
692    if (SUItr != sus.begin())
693      sus.erase(sus.begin(), SUItr);
694  }
695
696  // Remove all entries with empty su lists.
697  map.remove_if([&](std::pair<ValueType, SUList> &mapEntry) {
698      return (mapEntry.second.empty()); });
699
700  // Recompute the size of the map (NumNodes).
701  map.reComputeSize();
702}
703
704void ScheduleDAGInstrs::buildSchedGraph(AliasAnalysis *AA,
705                                        RegPressureTracker *RPTracker,
706                                        PressureDiffs *PDiffs,
707                                        LiveIntervals *LIS,
708                                        bool TrackLaneMasks) {
709  const TargetSubtargetInfo &ST = MF.getSubtarget();
710  bool UseAA = EnableAASchedMI.getNumOccurrences() > 0 ? EnableAASchedMI
711                                                       : ST.useAA();
712  AAForDep = UseAA ? AA : nullptr;
713
714  BarrierChain = nullptr;
715
716  this->TrackLaneMasks = TrackLaneMasks;
717  MISUnitMap.clear();
718  ScheduleDAG::clearDAG();
719
720  // Create an SUnit for each real instruction.
721  initSUnits();
722
723  if (PDiffs)
724    PDiffs->init(SUnits.size());
725
726  // We build scheduling units by walking a block's instruction list
727  // from bottom to top.
728
729  // Each MIs' memory operand(s) is analyzed to a list of underlying
730  // objects. The SU is then inserted in the SUList(s) mapped from the
731  // Value(s). Each Value thus gets mapped to lists of SUs depending
732  // on it, stores and loads kept separately. Two SUs are trivially
733  // non-aliasing if they both depend on only identified Values and do
734  // not share any common Value.
735  Value2SUsMap Stores, Loads(1 /*TrueMemOrderLatency*/);
736
737  // Certain memory accesses are known to not alias any SU in Stores
738  // or Loads, and have therefore their own 'NonAlias'
739  // domain. E.g. spill / reload instructions never alias LLVM I/R
740  // Values. It would be nice to assume that this type of memory
741  // accesses always have a proper memory operand modelling, and are
742  // therefore never unanalyzable, but this is conservatively not
743  // done.
744  Value2SUsMap NonAliasStores, NonAliasLoads(1 /*TrueMemOrderLatency*/);
745
746  // Track all instructions that may raise floating-point exceptions.
747  // These do not depend on one other (or normal loads or stores), but
748  // must not be rescheduled across global barriers.  Note that we don't
749  // really need a "map" here since we don't track those MIs by value;
750  // using the same Value2SUsMap data type here is simply a matter of
751  // convenience.
752  Value2SUsMap FPExceptions;
753
754  // Remove any stale debug info; sometimes BuildSchedGraph is called again
755  // without emitting the info from the previous call.
756  DbgValues.clear();
757  FirstDbgValue = nullptr;
758
759  assert(Defs.empty() && Uses.empty() &&
760         "Only BuildGraph should update Defs/Uses");
761  Defs.setUniverse(TRI->getNumRegs());
762  Uses.setUniverse(TRI->getNumRegs());
763
764  assert(CurrentVRegDefs.empty() && "nobody else should use CurrentVRegDefs");
765  assert(CurrentVRegUses.empty() && "nobody else should use CurrentVRegUses");
766  unsigned NumVirtRegs = MRI.getNumVirtRegs();
767  CurrentVRegDefs.setUniverse(NumVirtRegs);
768  CurrentVRegUses.setUniverse(NumVirtRegs);
769
770  // Model data dependencies between instructions being scheduled and the
771  // ExitSU.
772  addSchedBarrierDeps();
773
774  // Walk the list of instructions, from bottom moving up.
775  MachineInstr *DbgMI = nullptr;
776  for (MachineBasicBlock::iterator MII = RegionEnd, MIE = RegionBegin;
777       MII != MIE; --MII) {
778    MachineInstr &MI = *std::prev(MII);
779    if (DbgMI) {
780      DbgValues.push_back(std::make_pair(DbgMI, &MI));
781      DbgMI = nullptr;
782    }
783
784    if (MI.isDebugValue()) {
785      DbgMI = &MI;
786      continue;
787    }
788    if (MI.isDebugLabel())
789      continue;
790
791    SUnit *SU = MISUnitMap[&MI];
792    assert(SU && "No SUnit mapped to this MI");
793
794    if (RPTracker) {
795      RegisterOperands RegOpers;
796      RegOpers.collect(MI, *TRI, MRI, TrackLaneMasks, false);
797      if (TrackLaneMasks) {
798        SlotIndex SlotIdx = LIS->getInstructionIndex(MI);
799        RegOpers.adjustLaneLiveness(*LIS, MRI, SlotIdx);
800      }
801      if (PDiffs != nullptr)
802        PDiffs->addInstruction(SU->NodeNum, RegOpers, MRI);
803
804      if (RPTracker->getPos() == RegionEnd || &*RPTracker->getPos() != &MI)
805        RPTracker->recedeSkipDebugValues();
806      assert(&*RPTracker->getPos() == &MI && "RPTracker in sync");
807      RPTracker->recede(RegOpers);
808    }
809
810    assert(
811        (CanHandleTerminators || (!MI.isTerminator() && !MI.isPosition())) &&
812        "Cannot schedule terminators or labels!");
813
814    // Add register-based dependencies (data, anti, and output).
815    // For some instructions (calls, returns, inline-asm, etc.) there can
816    // be explicit uses and implicit defs, in which case the use will appear
817    // on the operand list before the def. Do two passes over the operand
818    // list to make sure that defs are processed before any uses.
819    bool HasVRegDef = false;
820    for (unsigned j = 0, n = MI.getNumOperands(); j != n; ++j) {
821      const MachineOperand &MO = MI.getOperand(j);
822      if (!MO.isReg() || !MO.isDef())
823        continue;
824      unsigned Reg = MO.getReg();
825      if (TargetRegisterInfo::isPhysicalRegister(Reg)) {
826        addPhysRegDeps(SU, j);
827      } else if (TargetRegisterInfo::isVirtualRegister(Reg)) {
828        HasVRegDef = true;
829        addVRegDefDeps(SU, j);
830      }
831    }
832    // Now process all uses.
833    for (unsigned j = 0, n = MI.getNumOperands(); j != n; ++j) {
834      const MachineOperand &MO = MI.getOperand(j);
835      // Only look at use operands.
836      // We do not need to check for MO.readsReg() here because subsequent
837      // subregister defs will get output dependence edges and need no
838      // additional use dependencies.
839      if (!MO.isReg() || !MO.isUse())
840        continue;
841      unsigned Reg = MO.getReg();
842      if (TargetRegisterInfo::isPhysicalRegister(Reg)) {
843        addPhysRegDeps(SU, j);
844      } else if (TargetRegisterInfo::isVirtualRegister(Reg) && MO.readsReg()) {
845        addVRegUseDeps(SU, j);
846      }
847    }
848
849    // If we haven't seen any uses in this scheduling region, create a
850    // dependence edge to ExitSU to model the live-out latency. This is required
851    // for vreg defs with no in-region use, and prefetches with no vreg def.
852    //
853    // FIXME: NumDataSuccs would be more precise than NumSuccs here. This
854    // check currently relies on being called before adding chain deps.
855    if (SU->NumSuccs == 0 && SU->Latency > 1 && (HasVRegDef || MI.mayLoad())) {
856      SDep Dep(SU, SDep::Artificial);
857      Dep.setLatency(SU->Latency - 1);
858      ExitSU.addPred(Dep);
859    }
860
861    // Add memory dependencies (Note: isStoreToStackSlot and
862    // isLoadFromStackSLot are not usable after stack slots are lowered to
863    // actual addresses).
864
865    // This is a barrier event that acts as a pivotal node in the DAG.
866    if (isGlobalMemoryObject(AA, &MI)) {
867
868      // Become the barrier chain.
869      if (BarrierChain)
870        BarrierChain->addPredBarrier(SU);
871      BarrierChain = SU;
872
873      LLVM_DEBUG(dbgs() << "Global memory object and new barrier chain: SU("
874                        << BarrierChain->NodeNum << ").\n";);
875
876      // Add dependencies against everything below it and clear maps.
877      addBarrierChain(Stores);
878      addBarrierChain(Loads);
879      addBarrierChain(NonAliasStores);
880      addBarrierChain(NonAliasLoads);
881      addBarrierChain(FPExceptions);
882
883      continue;
884    }
885
886    // Instructions that may raise FP exceptions may not be moved
887    // across any global barriers.
888    if (MI.mayRaiseFPException()) {
889      if (BarrierChain)
890        BarrierChain->addPredBarrier(SU);
891
892      FPExceptions.insert(SU, UnknownValue);
893
894      if (FPExceptions.size() >= HugeRegion) {
895        LLVM_DEBUG(dbgs() << "Reducing FPExceptions map.\n";);
896        Value2SUsMap empty;
897        reduceHugeMemNodeMaps(FPExceptions, empty, getReductionSize());
898      }
899    }
900
901    // If it's not a store or a variant load, we're done.
902    if (!MI.mayStore() &&
903        !(MI.mayLoad() && !MI.isDereferenceableInvariantLoad(AA)))
904      continue;
905
906    // Always add dependecy edge to BarrierChain if present.
907    if (BarrierChain)
908      BarrierChain->addPredBarrier(SU);
909
910    // Find the underlying objects for MI. The Objs vector is either
911    // empty, or filled with the Values of memory locations which this
912    // SU depends on.
913    UnderlyingObjectsVector Objs;
914    bool ObjsFound = getUnderlyingObjectsForInstr(&MI, MFI, Objs,
915                                                  MF.getDataLayout());
916
917    if (MI.mayStore()) {
918      if (!ObjsFound) {
919        // An unknown store depends on all stores and loads.
920        addChainDependencies(SU, Stores);
921        addChainDependencies(SU, NonAliasStores);
922        addChainDependencies(SU, Loads);
923        addChainDependencies(SU, NonAliasLoads);
924
925        // Map this store to 'UnknownValue'.
926        Stores.insert(SU, UnknownValue);
927      } else {
928        // Add precise dependencies against all previously seen memory
929        // accesses mapped to the same Value(s).
930        for (const UnderlyingObject &UnderlObj : Objs) {
931          ValueType V = UnderlObj.getValue();
932          bool ThisMayAlias = UnderlObj.mayAlias();
933
934          // Add dependencies to previous stores and loads mapped to V.
935          addChainDependencies(SU, (ThisMayAlias ? Stores : NonAliasStores), V);
936          addChainDependencies(SU, (ThisMayAlias ? Loads : NonAliasLoads), V);
937        }
938        // Update the store map after all chains have been added to avoid adding
939        // self-loop edge if multiple underlying objects are present.
940        for (const UnderlyingObject &UnderlObj : Objs) {
941          ValueType V = UnderlObj.getValue();
942          bool ThisMayAlias = UnderlObj.mayAlias();
943
944          // Map this store to V.
945          (ThisMayAlias ? Stores : NonAliasStores).insert(SU, V);
946        }
947        // The store may have dependencies to unanalyzable loads and
948        // stores.
949        addChainDependencies(SU, Loads, UnknownValue);
950        addChainDependencies(SU, Stores, UnknownValue);
951      }
952    } else { // SU is a load.
953      if (!ObjsFound) {
954        // An unknown load depends on all stores.
955        addChainDependencies(SU, Stores);
956        addChainDependencies(SU, NonAliasStores);
957
958        Loads.insert(SU, UnknownValue);
959      } else {
960        for (const UnderlyingObject &UnderlObj : Objs) {
961          ValueType V = UnderlObj.getValue();
962          bool ThisMayAlias = UnderlObj.mayAlias();
963
964          // Add precise dependencies against all previously seen stores
965          // mapping to the same Value(s).
966          addChainDependencies(SU, (ThisMayAlias ? Stores : NonAliasStores), V);
967
968          // Map this load to V.
969          (ThisMayAlias ? Loads : NonAliasLoads).insert(SU, V);
970        }
971        // The load may have dependencies to unanalyzable stores.
972        addChainDependencies(SU, Stores, UnknownValue);
973      }
974    }
975
976    // Reduce maps if they grow huge.
977    if (Stores.size() + Loads.size() >= HugeRegion) {
978      LLVM_DEBUG(dbgs() << "Reducing Stores and Loads maps.\n";);
979      reduceHugeMemNodeMaps(Stores, Loads, getReductionSize());
980    }
981    if (NonAliasStores.size() + NonAliasLoads.size() >= HugeRegion) {
982      LLVM_DEBUG(
983          dbgs() << "Reducing NonAliasStores and NonAliasLoads maps.\n";);
984      reduceHugeMemNodeMaps(NonAliasStores, NonAliasLoads, getReductionSize());
985    }
986  }
987
988  if (DbgMI)
989    FirstDbgValue = DbgMI;
990
991  Defs.clear();
992  Uses.clear();
993  CurrentVRegDefs.clear();
994  CurrentVRegUses.clear();
995
996  Topo.MarkDirty();
997}
998
999raw_ostream &llvm::operator<<(raw_ostream &OS, const PseudoSourceValue* PSV) {
1000  PSV->printCustom(OS);
1001  return OS;
1002}
1003
1004void ScheduleDAGInstrs::Value2SUsMap::dump() {
1005  for (auto &Itr : *this) {
1006    if (Itr.first.is<const Value*>()) {
1007      const Value *V = Itr.first.get<const Value*>();
1008      if (isa<UndefValue>(V))
1009        dbgs() << "Unknown";
1010      else
1011        V->printAsOperand(dbgs());
1012    }
1013    else if (Itr.first.is<const PseudoSourceValue*>())
1014      dbgs() <<  Itr.first.get<const PseudoSourceValue*>();
1015    else
1016      llvm_unreachable("Unknown Value type.");
1017
1018    dbgs() << " : ";
1019    dumpSUList(Itr.second);
1020  }
1021}
1022
1023void ScheduleDAGInstrs::reduceHugeMemNodeMaps(Value2SUsMap &stores,
1024                                              Value2SUsMap &loads, unsigned N) {
1025  LLVM_DEBUG(dbgs() << "Before reduction:\nStoring SUnits:\n"; stores.dump();
1026             dbgs() << "Loading SUnits:\n"; loads.dump());
1027
1028  // Insert all SU's NodeNums into a vector and sort it.
1029  std::vector<unsigned> NodeNums;
1030  NodeNums.reserve(stores.size() + loads.size());
1031  for (auto &I : stores)
1032    for (auto *SU : I.second)
1033      NodeNums.push_back(SU->NodeNum);
1034  for (auto &I : loads)
1035    for (auto *SU : I.second)
1036      NodeNums.push_back(SU->NodeNum);
1037  llvm::sort(NodeNums);
1038
1039  // The N last elements in NodeNums will be removed, and the SU with
1040  // the lowest NodeNum of them will become the new BarrierChain to
1041  // let the not yet seen SUs have a dependency to the removed SUs.
1042  assert(N <= NodeNums.size());
1043  SUnit *newBarrierChain = &SUnits[*(NodeNums.end() - N)];
1044  if (BarrierChain) {
1045    // The aliasing and non-aliasing maps reduce independently of each
1046    // other, but share a common BarrierChain. Check if the
1047    // newBarrierChain is above the former one. If it is not, it may
1048    // introduce a loop to use newBarrierChain, so keep the old one.
1049    if (newBarrierChain->NodeNum < BarrierChain->NodeNum) {
1050      BarrierChain->addPredBarrier(newBarrierChain);
1051      BarrierChain = newBarrierChain;
1052      LLVM_DEBUG(dbgs() << "Inserting new barrier chain: SU("
1053                        << BarrierChain->NodeNum << ").\n";);
1054    }
1055    else
1056      LLVM_DEBUG(dbgs() << "Keeping old barrier chain: SU("
1057                        << BarrierChain->NodeNum << ").\n";);
1058  }
1059  else
1060    BarrierChain = newBarrierChain;
1061
1062  insertBarrierChain(stores);
1063  insertBarrierChain(loads);
1064
1065  LLVM_DEBUG(dbgs() << "After reduction:\nStoring SUnits:\n"; stores.dump();
1066             dbgs() << "Loading SUnits:\n"; loads.dump());
1067}
1068
1069static void toggleKills(const MachineRegisterInfo &MRI, LivePhysRegs &LiveRegs,
1070                        MachineInstr &MI, bool addToLiveRegs) {
1071  for (MachineOperand &MO : MI.operands()) {
1072    if (!MO.isReg() || !MO.readsReg())
1073      continue;
1074    unsigned Reg = MO.getReg();
1075    if (!Reg)
1076      continue;
1077
1078    // Things that are available after the instruction are killed by it.
1079    bool IsKill = LiveRegs.available(MRI, Reg);
1080    MO.setIsKill(IsKill);
1081    if (addToLiveRegs)
1082      LiveRegs.addReg(Reg);
1083  }
1084}
1085
1086void ScheduleDAGInstrs::fixupKills(MachineBasicBlock &MBB) {
1087  LLVM_DEBUG(dbgs() << "Fixup kills for " << printMBBReference(MBB) << '\n');
1088
1089  LiveRegs.init(*TRI);
1090  LiveRegs.addLiveOuts(MBB);
1091
1092  // Examine block from end to start...
1093  for (MachineInstr &MI : make_range(MBB.rbegin(), MBB.rend())) {
1094    if (MI.isDebugInstr())
1095      continue;
1096
1097    // Update liveness.  Registers that are defed but not used in this
1098    // instruction are now dead. Mark register and all subregs as they
1099    // are completely defined.
1100    for (ConstMIBundleOperands O(MI); O.isValid(); ++O) {
1101      const MachineOperand &MO = *O;
1102      if (MO.isReg()) {
1103        if (!MO.isDef())
1104          continue;
1105        unsigned Reg = MO.getReg();
1106        if (!Reg)
1107          continue;
1108        LiveRegs.removeReg(Reg);
1109      } else if (MO.isRegMask()) {
1110        LiveRegs.removeRegsInMask(MO);
1111      }
1112    }
1113
1114    // If there is a bundle header fix it up first.
1115    if (!MI.isBundled()) {
1116      toggleKills(MRI, LiveRegs, MI, true);
1117    } else {
1118      MachineBasicBlock::instr_iterator Bundle = MI.getIterator();
1119      if (MI.isBundle())
1120        toggleKills(MRI, LiveRegs, MI, false);
1121
1122      // Some targets make the (questionable) assumtion that the instructions
1123      // inside the bundle are ordered and consequently only the last use of
1124      // a register inside the bundle can kill it.
1125      MachineBasicBlock::instr_iterator I = std::next(Bundle);
1126      while (I->isBundledWithSucc())
1127        ++I;
1128      do {
1129        if (!I->isDebugInstr())
1130          toggleKills(MRI, LiveRegs, *I, true);
1131        --I;
1132      } while (I != Bundle);
1133    }
1134  }
1135}
1136
1137void ScheduleDAGInstrs::dumpNode(const SUnit &SU) const {
1138#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
1139  dumpNodeName(SU);
1140  dbgs() << ": ";
1141  SU.getInstr()->dump();
1142#endif
1143}
1144
1145void ScheduleDAGInstrs::dump() const {
1146#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
1147  if (EntrySU.getInstr() != nullptr)
1148    dumpNodeAll(EntrySU);
1149  for (const SUnit &SU : SUnits)
1150    dumpNodeAll(SU);
1151  if (ExitSU.getInstr() != nullptr)
1152    dumpNodeAll(ExitSU);
1153#endif
1154}
1155
1156std::string ScheduleDAGInstrs::getGraphNodeLabel(const SUnit *SU) const {
1157  std::string s;
1158  raw_string_ostream oss(s);
1159  if (SU == &EntrySU)
1160    oss << "<entry>";
1161  else if (SU == &ExitSU)
1162    oss << "<exit>";
1163  else
1164    SU->getInstr()->print(oss, /*SkipOpers=*/true);
1165  return oss.str();
1166}
1167
1168/// Return the basic block label. It is not necessarilly unique because a block
1169/// contains multiple scheduling regions. But it is fine for visualization.
1170std::string ScheduleDAGInstrs::getDAGName() const {
1171  return "dag." + BB->getFullName();
1172}
1173
1174bool ScheduleDAGInstrs::canAddEdge(SUnit *SuccSU, SUnit *PredSU) {
1175  return SuccSU == &ExitSU || !Topo.IsReachable(PredSU, SuccSU);
1176}
1177
1178bool ScheduleDAGInstrs::addEdge(SUnit *SuccSU, const SDep &PredDep) {
1179  if (SuccSU != &ExitSU) {
1180    // Do not use WillCreateCycle, it assumes SD scheduling.
1181    // If Pred is reachable from Succ, then the edge creates a cycle.
1182    if (Topo.IsReachable(PredDep.getSUnit(), SuccSU))
1183      return false;
1184    Topo.AddPredQueued(SuccSU, PredDep.getSUnit());
1185  }
1186  SuccSU->addPred(PredDep, /*Required=*/!PredDep.isArtificial());
1187  // Return true regardless of whether a new edge needed to be inserted.
1188  return true;
1189}
1190
1191//===----------------------------------------------------------------------===//
1192// SchedDFSResult Implementation
1193//===----------------------------------------------------------------------===//
1194
1195namespace llvm {
1196
1197/// Internal state used to compute SchedDFSResult.
1198class SchedDFSImpl {
1199  SchedDFSResult &R;
1200
1201  /// Join DAG nodes into equivalence classes by their subtree.
1202  IntEqClasses SubtreeClasses;
1203  /// List PredSU, SuccSU pairs that represent data edges between subtrees.
1204  std::vector<std::pair<const SUnit *, const SUnit*>> ConnectionPairs;
1205
1206  struct RootData {
1207    unsigned NodeID;
1208    unsigned ParentNodeID;  ///< Parent node (member of the parent subtree).
1209    unsigned SubInstrCount = 0; ///< Instr count in this tree only, not
1210                                /// children.
1211
1212    RootData(unsigned id): NodeID(id),
1213                           ParentNodeID(SchedDFSResult::InvalidSubtreeID) {}
1214
1215    unsigned getSparseSetIndex() const { return NodeID; }
1216  };
1217
1218  SparseSet<RootData> RootSet;
1219
1220public:
1221  SchedDFSImpl(SchedDFSResult &r): R(r), SubtreeClasses(R.DFSNodeData.size()) {
1222    RootSet.setUniverse(R.DFSNodeData.size());
1223  }
1224
1225  /// Returns true if this node been visited by the DFS traversal.
1226  ///
1227  /// During visitPostorderNode the Node's SubtreeID is assigned to the Node
1228  /// ID. Later, SubtreeID is updated but remains valid.
1229  bool isVisited(const SUnit *SU) const {
1230    return R.DFSNodeData[SU->NodeNum].SubtreeID
1231      != SchedDFSResult::InvalidSubtreeID;
1232  }
1233
1234  /// Initializes this node's instruction count. We don't need to flag the node
1235  /// visited until visitPostorder because the DAG cannot have cycles.
1236  void visitPreorder(const SUnit *SU) {
1237    R.DFSNodeData[SU->NodeNum].InstrCount =
1238      SU->getInstr()->isTransient() ? 0 : 1;
1239  }
1240
1241  /// Called once for each node after all predecessors are visited. Revisit this
1242  /// node's predecessors and potentially join them now that we know the ILP of
1243  /// the other predecessors.
1244  void visitPostorderNode(const SUnit *SU) {
1245    // Mark this node as the root of a subtree. It may be joined with its
1246    // successors later.
1247    R.DFSNodeData[SU->NodeNum].SubtreeID = SU->NodeNum;
1248    RootData RData(SU->NodeNum);
1249    RData.SubInstrCount = SU->getInstr()->isTransient() ? 0 : 1;
1250
1251    // If any predecessors are still in their own subtree, they either cannot be
1252    // joined or are large enough to remain separate. If this parent node's
1253    // total instruction count is not greater than a child subtree by at least
1254    // the subtree limit, then try to join it now since splitting subtrees is
1255    // only useful if multiple high-pressure paths are possible.
1256    unsigned InstrCount = R.DFSNodeData[SU->NodeNum].InstrCount;
1257    for (const SDep &PredDep : SU->Preds) {
1258      if (PredDep.getKind() != SDep::Data)
1259        continue;
1260      unsigned PredNum = PredDep.getSUnit()->NodeNum;
1261      if ((InstrCount - R.DFSNodeData[PredNum].InstrCount) < R.SubtreeLimit)
1262        joinPredSubtree(PredDep, SU, /*CheckLimit=*/false);
1263
1264      // Either link or merge the TreeData entry from the child to the parent.
1265      if (R.DFSNodeData[PredNum].SubtreeID == PredNum) {
1266        // If the predecessor's parent is invalid, this is a tree edge and the
1267        // current node is the parent.
1268        if (RootSet[PredNum].ParentNodeID == SchedDFSResult::InvalidSubtreeID)
1269          RootSet[PredNum].ParentNodeID = SU->NodeNum;
1270      }
1271      else if (RootSet.count(PredNum)) {
1272        // The predecessor is not a root, but is still in the root set. This
1273        // must be the new parent that it was just joined to. Note that
1274        // RootSet[PredNum].ParentNodeID may either be invalid or may still be
1275        // set to the original parent.
1276        RData.SubInstrCount += RootSet[PredNum].SubInstrCount;
1277        RootSet.erase(PredNum);
1278      }
1279    }
1280    RootSet[SU->NodeNum] = RData;
1281  }
1282
1283  /// Called once for each tree edge after calling visitPostOrderNode on
1284  /// the predecessor. Increment the parent node's instruction count and
1285  /// preemptively join this subtree to its parent's if it is small enough.
1286  void visitPostorderEdge(const SDep &PredDep, const SUnit *Succ) {
1287    R.DFSNodeData[Succ->NodeNum].InstrCount
1288      += R.DFSNodeData[PredDep.getSUnit()->NodeNum].InstrCount;
1289    joinPredSubtree(PredDep, Succ);
1290  }
1291
1292  /// Adds a connection for cross edges.
1293  void visitCrossEdge(const SDep &PredDep, const SUnit *Succ) {
1294    ConnectionPairs.push_back(std::make_pair(PredDep.getSUnit(), Succ));
1295  }
1296
1297  /// Sets each node's subtree ID to the representative ID and record
1298  /// connections between trees.
1299  void finalize() {
1300    SubtreeClasses.compress();
1301    R.DFSTreeData.resize(SubtreeClasses.getNumClasses());
1302    assert(SubtreeClasses.getNumClasses() == RootSet.size()
1303           && "number of roots should match trees");
1304    for (const RootData &Root : RootSet) {
1305      unsigned TreeID = SubtreeClasses[Root.NodeID];
1306      if (Root.ParentNodeID != SchedDFSResult::InvalidSubtreeID)
1307        R.DFSTreeData[TreeID].ParentTreeID = SubtreeClasses[Root.ParentNodeID];
1308      R.DFSTreeData[TreeID].SubInstrCount = Root.SubInstrCount;
1309      // Note that SubInstrCount may be greater than InstrCount if we joined
1310      // subtrees across a cross edge. InstrCount will be attributed to the
1311      // original parent, while SubInstrCount will be attributed to the joined
1312      // parent.
1313    }
1314    R.SubtreeConnections.resize(SubtreeClasses.getNumClasses());
1315    R.SubtreeConnectLevels.resize(SubtreeClasses.getNumClasses());
1316    LLVM_DEBUG(dbgs() << R.getNumSubtrees() << " subtrees:\n");
1317    for (unsigned Idx = 0, End = R.DFSNodeData.size(); Idx != End; ++Idx) {
1318      R.DFSNodeData[Idx].SubtreeID = SubtreeClasses[Idx];
1319      LLVM_DEBUG(dbgs() << "  SU(" << Idx << ") in tree "
1320                        << R.DFSNodeData[Idx].SubtreeID << '\n');
1321    }
1322    for (const std::pair<const SUnit*, const SUnit*> &P : ConnectionPairs) {
1323      unsigned PredTree = SubtreeClasses[P.first->NodeNum];
1324      unsigned SuccTree = SubtreeClasses[P.second->NodeNum];
1325      if (PredTree == SuccTree)
1326        continue;
1327      unsigned Depth = P.first->getDepth();
1328      addConnection(PredTree, SuccTree, Depth);
1329      addConnection(SuccTree, PredTree, Depth);
1330    }
1331  }
1332
1333protected:
1334  /// Joins the predecessor subtree with the successor that is its DFS parent.
1335  /// Applies some heuristics before joining.
1336  bool joinPredSubtree(const SDep &PredDep, const SUnit *Succ,
1337                       bool CheckLimit = true) {
1338    assert(PredDep.getKind() == SDep::Data && "Subtrees are for data edges");
1339
1340    // Check if the predecessor is already joined.
1341    const SUnit *PredSU = PredDep.getSUnit();
1342    unsigned PredNum = PredSU->NodeNum;
1343    if (R.DFSNodeData[PredNum].SubtreeID != PredNum)
1344      return false;
1345
1346    // Four is the magic number of successors before a node is considered a
1347    // pinch point.
1348    unsigned NumDataSucs = 0;
1349    for (const SDep &SuccDep : PredSU->Succs) {
1350      if (SuccDep.getKind() == SDep::Data) {
1351        if (++NumDataSucs >= 4)
1352          return false;
1353      }
1354    }
1355    if (CheckLimit && R.DFSNodeData[PredNum].InstrCount > R.SubtreeLimit)
1356      return false;
1357    R.DFSNodeData[PredNum].SubtreeID = Succ->NodeNum;
1358    SubtreeClasses.join(Succ->NodeNum, PredNum);
1359    return true;
1360  }
1361
1362  /// Called by finalize() to record a connection between trees.
1363  void addConnection(unsigned FromTree, unsigned ToTree, unsigned Depth) {
1364    if (!Depth)
1365      return;
1366
1367    do {
1368      SmallVectorImpl<SchedDFSResult::Connection> &Connections =
1369        R.SubtreeConnections[FromTree];
1370      for (SchedDFSResult::Connection &C : Connections) {
1371        if (C.TreeID == ToTree) {
1372          C.Level = std::max(C.Level, Depth);
1373          return;
1374        }
1375      }
1376      Connections.push_back(SchedDFSResult::Connection(ToTree, Depth));
1377      FromTree = R.DFSTreeData[FromTree].ParentTreeID;
1378    } while (FromTree != SchedDFSResult::InvalidSubtreeID);
1379  }
1380};
1381
1382} // end namespace llvm
1383
1384namespace {
1385
1386/// Manage the stack used by a reverse depth-first search over the DAG.
1387class SchedDAGReverseDFS {
1388  std::vector<std::pair<const SUnit *, SUnit::const_pred_iterator>> DFSStack;
1389
1390public:
1391  bool isComplete() const { return DFSStack.empty(); }
1392
1393  void follow(const SUnit *SU) {
1394    DFSStack.push_back(std::make_pair(SU, SU->Preds.begin()));
1395  }
1396  void advance() { ++DFSStack.back().second; }
1397
1398  const SDep *backtrack() {
1399    DFSStack.pop_back();
1400    return DFSStack.empty() ? nullptr : std::prev(DFSStack.back().second);
1401  }
1402
1403  const SUnit *getCurr() const { return DFSStack.back().first; }
1404
1405  SUnit::const_pred_iterator getPred() const { return DFSStack.back().second; }
1406
1407  SUnit::const_pred_iterator getPredEnd() const {
1408    return getCurr()->Preds.end();
1409  }
1410};
1411
1412} // end anonymous namespace
1413
1414static bool hasDataSucc(const SUnit *SU) {
1415  for (const SDep &SuccDep : SU->Succs) {
1416    if (SuccDep.getKind() == SDep::Data &&
1417        !SuccDep.getSUnit()->isBoundaryNode())
1418      return true;
1419  }
1420  return false;
1421}
1422
1423/// Computes an ILP metric for all nodes in the subDAG reachable via depth-first
1424/// search from this root.
1425void SchedDFSResult::compute(ArrayRef<SUnit> SUnits) {
1426  if (!IsBottomUp)
1427    llvm_unreachable("Top-down ILP metric is unimplemented");
1428
1429  SchedDFSImpl Impl(*this);
1430  for (const SUnit &SU : SUnits) {
1431    if (Impl.isVisited(&SU) || hasDataSucc(&SU))
1432      continue;
1433
1434    SchedDAGReverseDFS DFS;
1435    Impl.visitPreorder(&SU);
1436    DFS.follow(&SU);
1437    while (true) {
1438      // Traverse the leftmost path as far as possible.
1439      while (DFS.getPred() != DFS.getPredEnd()) {
1440        const SDep &PredDep = *DFS.getPred();
1441        DFS.advance();
1442        // Ignore non-data edges.
1443        if (PredDep.getKind() != SDep::Data
1444            || PredDep.getSUnit()->isBoundaryNode()) {
1445          continue;
1446        }
1447        // An already visited edge is a cross edge, assuming an acyclic DAG.
1448        if (Impl.isVisited(PredDep.getSUnit())) {
1449          Impl.visitCrossEdge(PredDep, DFS.getCurr());
1450          continue;
1451        }
1452        Impl.visitPreorder(PredDep.getSUnit());
1453        DFS.follow(PredDep.getSUnit());
1454      }
1455      // Visit the top of the stack in postorder and backtrack.
1456      const SUnit *Child = DFS.getCurr();
1457      const SDep *PredDep = DFS.backtrack();
1458      Impl.visitPostorderNode(Child);
1459      if (PredDep)
1460        Impl.visitPostorderEdge(*PredDep, DFS.getCurr());
1461      if (DFS.isComplete())
1462        break;
1463    }
1464  }
1465  Impl.finalize();
1466}
1467
1468/// The root of the given SubtreeID was just scheduled. For all subtrees
1469/// connected to this tree, record the depth of the connection so that the
1470/// nearest connected subtrees can be prioritized.
1471void SchedDFSResult::scheduleTree(unsigned SubtreeID) {
1472  for (const Connection &C : SubtreeConnections[SubtreeID]) {
1473    SubtreeConnectLevels[C.TreeID] =
1474      std::max(SubtreeConnectLevels[C.TreeID], C.Level);
1475    LLVM_DEBUG(dbgs() << "  Tree: " << C.TreeID << " @"
1476                      << SubtreeConnectLevels[C.TreeID] << '\n');
1477  }
1478}
1479
1480#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
1481LLVM_DUMP_METHOD void ILPValue::print(raw_ostream &OS) const {
1482  OS << InstrCount << " / " << Length << " = ";
1483  if (!Length)
1484    OS << "BADILP";
1485  else
1486    OS << format("%g", ((double)InstrCount / Length));
1487}
1488
1489LLVM_DUMP_METHOD void ILPValue::dump() const {
1490  dbgs() << *this << '\n';
1491}
1492
1493namespace llvm {
1494
1495LLVM_DUMP_METHOD
1496raw_ostream &operator<<(raw_ostream &OS, const ILPValue &Val) {
1497  Val.print(OS);
1498  return OS;
1499}
1500
1501} // end namespace llvm
1502
1503#endif
1504