C++ BranchInst类代码示例

void vSSA::createSigmasIfNeeded(BasicBlock *BB)
{
	TerminatorInst *ti = BB->getTerminator();
	// If the condition used in the terminator instruction is a Comparison instruction:
	//for each operand of the CmpInst, create sigmas, depending on some conditions
	/*
	if(isa<BranchInst>(ti)){
		BranchInst * bc = cast<BranchInst>(ti);
		if(bc->isConditional()){
			Value * cond = bc->getCondition();
			CmpInst *comparison = dyn_cast<CmpInst>(cond);
			for (User::const_op_iterator it = comparison->op_begin(), e = comparison->op_end(); it != e; ++it) {
				Value *operand = *it;
				if (isa<Instruction>(operand) || isa<Argument>(operand)) {
					insertSigmas(ti, operand);
				}
			}
		}
	}
	*/
	
	// CASE 1: Branch Instruction
	BranchInst *bi = NULL;
	SwitchInst *si = NULL;
	if ((bi = dyn_cast<BranchInst>(ti))) {
		if (bi->isConditional()) {
			Value *condition = bi->getCondition();
			
			ICmpInst *comparison = dyn_cast<ICmpInst>(condition);
			
			if (comparison) {
				// Create sigmas for ICmp operands
				for (User::const_op_iterator opit = comparison->op_begin(), opend = comparison->op_end(); opit != opend; ++opit) {
					Value *operand = *opit;
					
					if (isa<Instruction>(operand) || isa<Argument>(operand)) {
						insertSigmas(ti, operand);
						
						// If the operand is a result of a indirect instruction (e.g. ZExt, SExt, Trunc),
						// Create sigmas for the operands of the operands too
						CastInst *cinst = NULL;
						if ((cinst = dyn_cast<CastInst>(operand))) {
							insertSigmas(ti, cinst->getOperand(0));
						}
					}
				}
			}
		}
	}
	// CASE 2: Switch Instruction
	
	else if ((si = dyn_cast<SwitchInst>(ti))) {
		Value *condition = si->getCondition();
		
		if (isa<Instruction>(condition) || isa<Argument>(condition)) {
			insertSigmas(ti, condition);
			
			// If the operand is a result of a indirect instruction (e.g. ZExt, SExt, Trunc),
			// Create sigmas for the operands of the operands too
			CastInst *cinst = NULL;
			if ((cinst = dyn_cast<CastInst>(condition))) {
				insertSigmas(ti, cinst->getOperand(0));
			}
		}
	}
	
}

void CPFlowFunction::visitBranchInst(BranchInst &BI) {
  CPLatticePoint* result = new CPLatticePoint(*(info_in_casted.back()));
  info_in_casted.pop_back();
  BranchInst* current = &BI;

  if (BI.isConditional()) {
    Value* cond = BI.getCondition();
    if (isa<ICmpInst>(cond)) {
      std::pair<Use*, Use *> branches = helper::getOps(BI);
      Use* true_branch = branches.first;
      Use* false_branch = branches.second;

      ICmpInst* cmp = dyn_cast<ICmpInst>(cond);
      std::pair<Use*, Use *> operands = helper::getOps(*cmp);
      Use* rhs = operands.second;
      Use* lhs = operands.first;

      ConstantInt* rhs_const = NULL;
      ConstantInt* lhs_const = NULL;
      // get the rhs/lhs as a constant int
      if (isa<ConstantInt>(rhs)) {
        rhs_const = dyn_cast<ConstantInt>(rhs);
      } else if (result->representation.count(rhs->get()) > 0) {
        rhs_const = result->representation[rhs->get()];
      } else {
        rhs_const = ConstantInt::get(context, llvm::APInt(32, 0, true));
      } 
      if (isa<ConstantInt>(lhs)) {
        lhs_const = dyn_cast<ConstantInt>(lhs->get());
      } else if (result->representation.count(lhs->get()) > 0) {
        lhs_const = result->representation[lhs->get()];
      } else {
        lhs_const = ConstantInt::get(context, llvm::APInt(32, 0, true));
      }

      // Create successors
      CPLatticePoint* true_branchCLP = new CPLatticePoint(false, false, std::map<Value*,ConstantInt*>(result->representation));
      CPLatticePoint* false_branchCLP = new CPLatticePoint(false, false, std::map<Value*,ConstantInt*>(result->representation));

      // get the predicate
      int predicate = 0;
      predicate = cmp->isSigned() ? cmp->getSignedPredicate() : cmp->getUnsignedPredicate();
      if (predicate == CmpInst::ICMP_EQ) {
        if (isa<ConstantInt>(lhs)) {
           true_branchCLP->representation[rhs->get()] = lhs_const;
        } else if (isa<ConstantInt>(rhs)) {
           true_branchCLP->representation[lhs->get()] = rhs_const;
        }
        out_map[true_branch->get()] = true_branchCLP;
        out_map[false_branch->get()] = false_branchCLP;
      } else if (predicate == CmpInst::ICMP_NE) {
        if (isa<ConstantInt>(lhs)) {
           false_branchCLP->representation[rhs->get()] = lhs_const;
        } else if (isa<ConstantInt>(rhs)) {
           false_branchCLP->representation[lhs->get()] = rhs_const;
        }
        out_map[true_branch->get()] = true_branchCLP;
        out_map[false_branch->get()] = false_branchCLP;
      } else {
        for (std::map<Value *, LatticePoint *>::iterator it=out_map.begin(); it != out_map.end(); ++it){
          Value* elm = it->first;
          out_map[elm] = new CPLatticePoint(*result);
        }
      }
    } else {
      for (std::map<Value *, LatticePoint *>::iterator it=out_map.begin(); it != out_map.end(); ++it){
        Value* elm = it->first;
        out_map[elm] = new CPLatticePoint(*result);
      }
    }
  } else {
    for (std::map<Value *, LatticePoint *>::iterator it=out_map.begin(); it != out_map.end(); ++it){
        Value* elm = it->first;
        out_map[elm] = new CPLatticePoint(*result);
    }
  }
}

std::pair<Optional<SILValue>, SILLocation>
SILGenFunction::emitEpilogBB(SILLocation TopLevel) {
  assert(ReturnDest.getBlock() && "no epilog bb prepared?!");
  SILBasicBlock *epilogBB = ReturnDest.getBlock();
  SILLocation ImplicitReturnFromTopLevel =
    ImplicitReturnLocation::getImplicitReturnLoc(TopLevel);
  SmallVector<SILValue, 4> directResults;
  Optional<SILLocation> returnLoc = None;

  // If the current BB isn't terminated, and we require a return, then we
  // are not allowed to fall off the end of the function and can't reach here.
  if (NeedsReturn && B.hasValidInsertionPoint())
    B.createUnreachable(ImplicitReturnFromTopLevel);

  if (epilogBB->pred_empty()) {
    // If the epilog was not branched to at all, kill the BB and
    // just emit the epilog into the current BB.
    while (!epilogBB->empty())
      epilogBB->back().eraseFromParent();
    eraseBasicBlock(epilogBB);

    // If the current bb is terminated then the epilog is just unreachable.
    if (!B.hasValidInsertionPoint())
      return { None, TopLevel };

    // We emit the epilog at the current insertion point.
    returnLoc = ImplicitReturnFromTopLevel;

  } else if (std::next(epilogBB->pred_begin()) == epilogBB->pred_end()
             && !B.hasValidInsertionPoint()) {
    // If the epilog has a single predecessor and there's no current insertion
    // point to fall through from, then we can weld the epilog to that
    // predecessor BB.

    // Steal the branch argument as the return value if present.
    SILBasicBlock *pred = *epilogBB->pred_begin();
    BranchInst *predBranch = cast<BranchInst>(pred->getTerminator());
    assert(predBranch->getArgs().size() == epilogBB->bbarg_size() &&
           "epilog predecessor arguments does not match block params");

    for (auto index : indices(predBranch->getArgs())) {
      SILValue result = predBranch->getArgs()[index];
      directResults.push_back(result);
      epilogBB->getBBArg(index)->replaceAllUsesWith(result);
    }

    // If we are optimizing, we should use the return location from the single,
    // previously processed, return statement if any.
    if (predBranch->getLoc().is<ReturnLocation>()) {
      returnLoc = predBranch->getLoc();
    } else {
      returnLoc = ImplicitReturnFromTopLevel;
    }
    
    // Kill the branch to the now-dead epilog BB.
    pred->erase(predBranch);

    // Move any instructions from the EpilogBB to the end of the 'pred' block.
    pred->spliceAtEnd(epilogBB);

    // Finally we can erase the epilog BB.
    eraseBasicBlock(epilogBB);

    // Emit the epilog into its former predecessor.
    B.setInsertionPoint(pred);
  } else {
    // Move the epilog block to the end of the ordinary section.
    auto endOfOrdinarySection = StartOfPostmatter;
    B.moveBlockTo(epilogBB, endOfOrdinarySection);

    // Emit the epilog into the epilog bb. Its arguments are the
    // direct results.
    directResults.append(epilogBB->bbarg_begin(), epilogBB->bbarg_end());

    // If we are falling through from the current block, the return is implicit.
    B.emitBlock(epilogBB, ImplicitReturnFromTopLevel);
  }
  
  // Emit top-level cleanups into the epilog block.
  assert(!Cleanups.hasAnyActiveCleanups(getCleanupsDepth(),
                                        ReturnDest.getDepth()) &&
         "emitting epilog in wrong scope");

  auto cleanupLoc = CleanupLocation::get(TopLevel);
  Cleanups.emitCleanupsForReturn(cleanupLoc);

  // If the return location is known to be that of an already
  // processed return, use it. (This will get triggered when the
  // epilog logic is simplified.)
  //
  // Otherwise make the ret instruction part of the cleanups.
  if (!returnLoc) returnLoc = cleanupLoc;

  // Build the return value.  We don't do this if there are no direct
  // results; this can happen for void functions, but also happens when
  // prepareEpilog was asked to not add result arguments to the epilog
  // block.
  SILValue returnValue;
  if (!directResults.empty()) {
    assert(directResults.size()
//.........这里部分代码省略.........

/// Insert code in the prolog code when unrolling a loop with a
/// run-time trip-count.
///
/// This method assumes that the loop unroll factor is total number
/// of loop bodes in the loop after unrolling. (Some folks refer
/// to the unroll factor as the number of *extra* copies added).
/// We assume also that the loop unroll factor is a power-of-two. So, after
/// unrolling the loop, the number of loop bodies executed is 2,
/// 4, 8, etc.  Note - LLVM converts the if-then-sequence to a switch
/// instruction in SimplifyCFG.cpp.  Then, the backend decides how code for
/// the switch instruction is generated.
///
///    extraiters = tripcount % loopfactor
///    if (extraiters == 0) jump Loop:
///    if (extraiters == loopfactor) jump L1
///    if (extraiters == loopfactor-1) jump L2
///    ...
///    L1:  LoopBody;
///    L2:  LoopBody;
///    ...
///    if tripcount < loopfactor jump End
///    Loop:
///    ...
///    End:
///
bool llvm::UnrollRuntimeLoopProlog(Loop *L, unsigned Count, LoopInfo *LI,
                                   LPPassManager *LPM) {
  // for now, only unroll loops that contain a single exit
  if (!L->getExitingBlock())
    return false;

  // Make sure the loop is in canonical form, and there is a single
  // exit block only.
  if (!L->isLoopSimplifyForm() || L->getUniqueExitBlock() == 0)
    return false;

  // Use Scalar Evolution to compute the trip count.  This allows more
  // loops to be unrolled than relying on induction var simplification
  if (!LPM)
    return false;
  ScalarEvolution *SE = LPM->getAnalysisIfAvailable<ScalarEvolution>();
  if (SE == 0)
    return false;

  // Only unroll loops with a computable trip count and the trip count needs
  // to be an int value (allowing a pointer type is a TODO item)
  const SCEV *BECount = SE->getBackedgeTakenCount(L);
  if (isa<SCEVCouldNotCompute>(BECount) || !BECount->getType()->isIntegerTy())
    return false;

  // Add 1 since the backedge count doesn't include the first loop iteration
  const SCEV *TripCountSC =
    SE->getAddExpr(BECount, SE->getConstant(BECount->getType(), 1));
  if (isa<SCEVCouldNotCompute>(TripCountSC))
    return false;

  // We only handle cases when the unroll factor is a power of 2.
  // Count is the loop unroll factor, the number of extra copies added + 1.
  if ((Count & (Count-1)) != 0)
    return false;

  // If this loop is nested, then the loop unroller changes the code in
  // parent loop, so the Scalar Evolution pass needs to be run again
  if (Loop *ParentLoop = L->getParentLoop())
    SE->forgetLoop(ParentLoop);

  BasicBlock *PH = L->getLoopPreheader();
  BasicBlock *Header = L->getHeader();
  BasicBlock *Latch = L->getLoopLatch();
  // It helps to splits the original preheader twice, one for the end of the
  // prolog code and one for a new loop preheader
  BasicBlock *PEnd = SplitEdge(PH, Header, LPM->getAsPass());
  BasicBlock *NewPH = SplitBlock(PEnd, PEnd->getTerminator(), LPM->getAsPass());
  BranchInst *PreHeaderBR = cast<BranchInst>(PH->getTerminator());

  // Compute the number of extra iterations required, which is:
  //  extra iterations = run-time trip count % (loop unroll factor + 1)
  SCEVExpander Expander(*SE, "loop-unroll");
  Value *TripCount = Expander.expandCodeFor(TripCountSC, TripCountSC->getType(),
                                            PreHeaderBR);
  Type *CountTy = TripCount->getType();
  BinaryOperator *ModVal =
    BinaryOperator::CreateURem(TripCount,
                               ConstantInt::get(CountTy, Count),
                               "xtraiter");
  ModVal->insertBefore(PreHeaderBR);

  // Check if for no extra iterations, then jump to unrolled loop
  Value *BranchVal = new ICmpInst(PreHeaderBR,
                                  ICmpInst::ICMP_NE, ModVal,
                                  ConstantInt::get(CountTy, 0), "lcmp");
  // Branch to either the extra iterations or the unrolled loop
  // We will fix up the true branch label when adding loop body copies
  BranchInst::Create(PEnd, PEnd, BranchVal, PreHeaderBR);
  assert(PreHeaderBR->isUnconditional() &&
         PreHeaderBR->getSuccessor(0) == PEnd &&
         "CFG edges in Preheader are not correct");
  PreHeaderBR->eraseFromParent();

  ValueToValueMapTy LVMap;
//.........这里部分代码省略.........

bool LoopInterchangeTransform::adjustLoopBranches() {

  DEBUG(dbgs() << "adjustLoopBranches called\n");
  // Adjust the loop preheader
  BasicBlock *InnerLoopHeader = InnerLoop->getHeader();
  BasicBlock *OuterLoopHeader = OuterLoop->getHeader();
  BasicBlock *InnerLoopLatch = InnerLoop->getLoopLatch();
  BasicBlock *OuterLoopLatch = OuterLoop->getLoopLatch();
  BasicBlock *OuterLoopPreHeader = OuterLoop->getLoopPreheader();
  BasicBlock *InnerLoopPreHeader = InnerLoop->getLoopPreheader();
  BasicBlock *OuterLoopPredecessor = OuterLoopPreHeader->getUniquePredecessor();
  BasicBlock *InnerLoopLatchPredecessor =
      InnerLoopLatch->getUniquePredecessor();
  BasicBlock *InnerLoopLatchSuccessor;
  BasicBlock *OuterLoopLatchSuccessor;

  BranchInst *OuterLoopLatchBI =
      dyn_cast<BranchInst>(OuterLoopLatch->getTerminator());
  BranchInst *InnerLoopLatchBI =
      dyn_cast<BranchInst>(InnerLoopLatch->getTerminator());
  BranchInst *OuterLoopHeaderBI =
      dyn_cast<BranchInst>(OuterLoopHeader->getTerminator());
  BranchInst *InnerLoopHeaderBI =
      dyn_cast<BranchInst>(InnerLoopHeader->getTerminator());

  if (!OuterLoopPredecessor || !InnerLoopLatchPredecessor ||
      !OuterLoopLatchBI || !InnerLoopLatchBI || !OuterLoopHeaderBI ||
      !InnerLoopHeaderBI)
    return false;

  BranchInst *InnerLoopLatchPredecessorBI =
      dyn_cast<BranchInst>(InnerLoopLatchPredecessor->getTerminator());
  BranchInst *OuterLoopPredecessorBI =
      dyn_cast<BranchInst>(OuterLoopPredecessor->getTerminator());

  if (!OuterLoopPredecessorBI || !InnerLoopLatchPredecessorBI)
    return false;
  BasicBlock *InnerLoopHeaderSuccessor = InnerLoopHeader->getUniqueSuccessor();
  if (!InnerLoopHeaderSuccessor)
    return false;

  // Adjust Loop Preheader and headers

  unsigned NumSucc = OuterLoopPredecessorBI->getNumSuccessors();
  for (unsigned i = 0; i < NumSucc; ++i) {
    if (OuterLoopPredecessorBI->getSuccessor(i) == OuterLoopPreHeader)
      OuterLoopPredecessorBI->setSuccessor(i, InnerLoopPreHeader);
  }

  NumSucc = OuterLoopHeaderBI->getNumSuccessors();
  for (unsigned i = 0; i < NumSucc; ++i) {
    if (OuterLoopHeaderBI->getSuccessor(i) == OuterLoopLatch)
      OuterLoopHeaderBI->setSuccessor(i, LoopExit);
    else if (OuterLoopHeaderBI->getSuccessor(i) == InnerLoopPreHeader)
      OuterLoopHeaderBI->setSuccessor(i, InnerLoopHeaderSuccessor);
  }

  // Adjust reduction PHI's now that the incoming block has changed.
  updateIncomingBlock(InnerLoopHeaderSuccessor, InnerLoopHeader,
                      OuterLoopHeader);

  BranchInst::Create(OuterLoopPreHeader, InnerLoopHeaderBI);
  InnerLoopHeaderBI->eraseFromParent();

  // -------------Adjust loop latches-----------
  if (InnerLoopLatchBI->getSuccessor(0) == InnerLoopHeader)
    InnerLoopLatchSuccessor = InnerLoopLatchBI->getSuccessor(1);
  else
    InnerLoopLatchSuccessor = InnerLoopLatchBI->getSuccessor(0);

  NumSucc = InnerLoopLatchPredecessorBI->getNumSuccessors();
  for (unsigned i = 0; i < NumSucc; ++i) {
    if (InnerLoopLatchPredecessorBI->getSuccessor(i) == InnerLoopLatch)
      InnerLoopLatchPredecessorBI->setSuccessor(i, InnerLoopLatchSuccessor);
  }

  // Adjust PHI nodes in InnerLoopLatchSuccessor. Update all uses of PHI with
  // the value and remove this PHI node from inner loop.
  SmallVector<PHINode *, 8> LcssaVec;
  for (auto I = InnerLoopLatchSuccessor->begin(); isa<PHINode>(I); ++I) {
    PHINode *LcssaPhi = cast<PHINode>(I);
    LcssaVec.push_back(LcssaPhi);
  }
  for (auto I = LcssaVec.begin(), E = LcssaVec.end(); I != E; ++I) {
    PHINode *P = *I;
    Value *Incoming = P->getIncomingValueForBlock(InnerLoopLatch);
    P->replaceAllUsesWith(Incoming);
    P->eraseFromParent();
  }

  if (OuterLoopLatchBI->getSuccessor(0) == OuterLoopHeader)
    OuterLoopLatchSuccessor = OuterLoopLatchBI->getSuccessor(1);
  else
    OuterLoopLatchSuccessor = OuterLoopLatchBI->getSuccessor(0);

  if (InnerLoopLatchBI->getSuccessor(1) == InnerLoopLatchSuccessor)
    InnerLoopLatchBI->setSuccessor(1, OuterLoopLatchSuccessor);
  else
    InnerLoopLatchBI->setSuccessor(0, OuterLoopLatchSuccessor);

//.........这里部分代码省略.........

/// \brief Simplify one loop and queue further loops for simplification.
///
/// FIXME: Currently this accepts both lots of analyses that it uses and a raw
/// Pass pointer. The Pass pointer is used by numerous utilities to update
/// specific analyses. Rather than a pass it would be much cleaner and more
/// explicit if they accepted the analysis directly and then updated it.
static bool simplifyOneLoop(Loop *L, SmallVectorImpl<Loop *> &Worklist,
                            AliasAnalysis *AA, DominatorTree *DT, LoopInfo *LI,
                            ScalarEvolution *SE, Pass *PP,
                            AssumptionCache *AC) {
  bool Changed = false;
ReprocessLoop:

  // Check to see that no blocks (other than the header) in this loop have
  // predecessors that are not in the loop.  This is not valid for natural
  // loops, but can occur if the blocks are unreachable.  Since they are
  // unreachable we can just shamelessly delete those CFG edges!
  for (Loop::block_iterator BB = L->block_begin(), E = L->block_end();
       BB != E; ++BB) {
    if (*BB == L->getHeader()) continue;

    SmallPtrSet<BasicBlock*, 4> BadPreds;
    for (pred_iterator PI = pred_begin(*BB),
         PE = pred_end(*BB); PI != PE; ++PI) {
      BasicBlock *P = *PI;
      if (!L->contains(P))
        BadPreds.insert(P);
    }

    // Delete each unique out-of-loop (and thus dead) predecessor.
    for (BasicBlock *P : BadPreds) {

      DEBUG(dbgs() << "LoopSimplify: Deleting edge from dead predecessor "
                   << P->getName() << "\n");

      // Inform each successor of each dead pred.
      for (succ_iterator SI = succ_begin(P), SE = succ_end(P); SI != SE; ++SI)
        (*SI)->removePredecessor(P);
      // Zap the dead pred's terminator and replace it with unreachable.
      TerminatorInst *TI = P->getTerminator();
       TI->replaceAllUsesWith(UndefValue::get(TI->getType()));
      P->getTerminator()->eraseFromParent();
      new UnreachableInst(P->getContext(), P);
      Changed = true;
    }
  }

  // If there are exiting blocks with branches on undef, resolve the undef in
  // the direction which will exit the loop. This will help simplify loop
  // trip count computations.
  SmallVector<BasicBlock*, 8> ExitingBlocks;
  L->getExitingBlocks(ExitingBlocks);
  for (SmallVectorImpl<BasicBlock *>::iterator I = ExitingBlocks.begin(),
       E = ExitingBlocks.end(); I != E; ++I)
    if (BranchInst *BI = dyn_cast<BranchInst>((*I)->getTerminator()))
      if (BI->isConditional()) {
        if (UndefValue *Cond = dyn_cast<UndefValue>(BI->getCondition())) {

          DEBUG(dbgs() << "LoopSimplify: Resolving \"br i1 undef\" to exit in "
                       << (*I)->getName() << "\n");

          BI->setCondition(ConstantInt::get(Cond->getType(),
                                            !L->contains(BI->getSuccessor(0))));

          // This may make the loop analyzable, force SCEV recomputation.
          if (SE)
            SE->forgetLoop(L);

          Changed = true;
        }
      }

  // Does the loop already have a preheader?  If so, don't insert one.
  BasicBlock *Preheader = L->getLoopPreheader();
  if (!Preheader) {
    Preheader = InsertPreheaderForLoop(L, PP);
    if (Preheader) {
      ++NumInserted;
      Changed = true;
    }
  }

  // Next, check to make sure that all exit nodes of the loop only have
  // predecessors that are inside of the loop.  This check guarantees that the
  // loop preheader/header will dominate the exit blocks.  If the exit block has
  // predecessors from outside of the loop, split the edge now.
  SmallVector<BasicBlock*, 8> ExitBlocks;
  L->getExitBlocks(ExitBlocks);

  SmallSetVector<BasicBlock *, 8> ExitBlockSet(ExitBlocks.begin(),
                                               ExitBlocks.end());
  for (SmallSetVector<BasicBlock *, 8>::iterator I = ExitBlockSet.begin(),
         E = ExitBlockSet.end(); I != E; ++I) {
    BasicBlock *ExitBlock = *I;
    for (pred_iterator PI = pred_begin(ExitBlock), PE = pred_end(ExitBlock);
         PI != PE; ++PI)
      // Must be exactly this loop: no subloops, parent loops, or non-loop preds
      // allowed.
      if (!L->contains(*PI)) {
        if (rewriteLoopExitBlock(L, ExitBlock, AA, DT, LI, PP)) {
//.........这里部分代码省略.........

//.........这里部分代码省略.........
      for (; (PN = dyn_cast<PHINode>(I)); ++I) {
        for (std::map<BasicBlock*, unsigned>::iterator PCI =PredCount.begin(),
             E = PredCount.end(); PCI != E; ++PCI) {
          BasicBlock *Pred     = PCI->first;
          for (unsigned NumToRemove = PCI->second; NumToRemove; --NumToRemove)
            PN->removeIncomingValue(Pred, false);
        }
      }
    }
    
    // If the loops above have made these phi nodes have 0 or 1 operand,
    // replace them with undef or the input value.  We must do this for
    // correctness, because 0-operand phis are not valid.
    PN = cast<PHINode>(NewBB->begin());
    if (PN->getNumIncomingValues() == 0) {
      BasicBlock::iterator I = NewBB->begin();
      BasicBlock::const_iterator OldI = OldBB->begin();
      while ((PN = dyn_cast<PHINode>(I++))) {
        Value *NV = UndefValue::get(PN->getType());
        PN->replaceAllUsesWith(NV);
        assert(VMap[&*OldI] == PN && "VMap mismatch");
        VMap[&*OldI] = NV;
        PN->eraseFromParent();
        ++OldI;
      }
    }
  }

  // Make a second pass over the PHINodes now that all of them have been
  // remapped into the new function, simplifying the PHINode and performing any
  // recursive simplifications exposed. This will transparently update the
  // WeakVH in the VMap. Notably, we rely on that so that if we coalesce
  // two PHINodes, the iteration over the old PHIs remains valid, and the
  // mapping will just map us to the new node (which may not even be a PHI
  // node).
  for (unsigned Idx = 0, Size = PHIToResolve.size(); Idx != Size; ++Idx)
    if (PHINode *PN = dyn_cast<PHINode>(VMap[PHIToResolve[Idx]]))
      recursivelySimplifyInstruction(PN);

  // Now that the inlined function body has been fully constructed, go through
  // and zap unconditional fall-through branches. This happens all the time when
  // specializing code: code specialization turns conditional branches into
  // uncond branches, and this code folds them.
  Function::iterator Begin = cast<BasicBlock>(VMap[StartingBB])->getIterator();
  Function::iterator I = Begin;
  while (I != NewFunc->end()) {
    // Check if this block has become dead during inlining or other
    // simplifications. Note that the first block will appear dead, as it has
    // not yet been wired up properly.
    if (I != Begin && (pred_begin(&*I) == pred_end(&*I) ||
                       I->getSinglePredecessor() == &*I)) {
      BasicBlock *DeadBB = &*I++;
      DeleteDeadBlock(DeadBB);
      continue;
    }

    // We need to simplify conditional branches and switches with a constant
    // operand. We try to prune these out when cloning, but if the
    // simplification required looking through PHI nodes, those are only
    // available after forming the full basic block. That may leave some here,
    // and we still want to prune the dead code as early as possible.
    ConstantFoldTerminator(&*I);

    BranchInst *BI = dyn_cast<BranchInst>(I->getTerminator());
    if (!BI || BI->isConditional()) { ++I; continue; }
    
    BasicBlock *Dest = BI->getSuccessor(0);
    if (!Dest->getSinglePredecessor()) {
      ++I; continue;
    }

    // We shouldn't be able to get single-entry PHI nodes here, as instsimplify
    // above should have zapped all of them..
    assert(!isa<PHINode>(Dest->begin()));

    // We know all single-entry PHI nodes in the inlined function have been
    // removed, so we just need to splice the blocks.
    BI->eraseFromParent();
    
    // Make all PHI nodes that referred to Dest now refer to I as their source.
    Dest->replaceAllUsesWith(&*I);

    // Move all the instructions in the succ to the pred.
    I->getInstList().splice(I->end(), Dest->getInstList());
    
    // Remove the dest block.
    Dest->eraseFromParent();
    
    // Do not increment I, iteratively merge all things this block branches to.
  }

  // Make a final pass over the basic blocks from the old function to gather
  // any return instructions which survived folding. We have to do this here
  // because we can iteratively remove and merge returns above.
  for (Function::iterator I = cast<BasicBlock>(VMap[StartingBB])->getIterator(),
                          E = NewFunc->end();
       I != E; ++I)
    if (ReturnInst *RI = dyn_cast<ReturnInst>(I->getTerminator()))
      Returns.push_back(RI);
}

/// If \param [in] BB has more than one predecessor that is a conditional
/// branch, attempt to use parallel and/or for the branch condition. \returns
/// true on success.
///
/// Before:
///   ......
///   %cmp10 = fcmp une float %tmp1, %tmp2
///   br i1 %cmp1, label %if.then, label %lor.rhs
///
/// lor.rhs:
///   ......
///   %cmp11 = fcmp une float %tmp3, %tmp4
///   br i1 %cmp11, label %if.then, label %ifend
///
/// if.end:  // the merge block
///   ......
///
/// if.then: // has two predecessors, both of them contains conditional branch.
///   ......
///   br label %if.end;
///
/// After:
///  ......
///  %cmp10 = fcmp une float %tmp1, %tmp2
///  ......
///  %cmp11 = fcmp une float %tmp3, %tmp4
///  %cmp12 = or i1 %cmp10, %cmp11    // parallel-or mode.
///  br i1 %cmp12, label %if.then, label %ifend
///
///  if.end:
///    ......
///
///  if.then:
///    ......
///    br label %if.end;
///
///  Current implementation handles two cases.
///  Case 1: \param BB is on the else-path.
///
///          BB1
///        /     |
///       BB2    |
///      /   \   |
///     BB3   \  |     where, BB1, BB2 contain conditional branches.
///      \    |  /     BB3 contains unconditional branch.
///       \   | /      BB4 corresponds to \param BB which is also the merge.
///  BB => BB4
///
///
///  Corresponding source code:
///
///  if (a == b && c == d)
///    statement; // BB3
///
///  Case 2: \param BB BB is on the then-path.
///
///             BB1
///          /      |
///         |      BB2
///         \    /    |  where BB1, BB2 contain conditional branches.
///  BB =>   BB3      |  BB3 contains unconditiona branch and corresponds
///           \     /    to \param BB.  BB4 is the merge.
///             BB4
///
///  Corresponding source code:
///
///  if (a == b || c == d)
///    statement;  // BB3
///
///  In both cases,  \param BB is the common successor of conditional branches.
///  In Case 1, \param BB (BB4) has an unconditional branch (BB3) as
///  its predecessor.  In Case 2, \param BB (BB3) only has conditional branches
///  as its predecessors.
///
bool FlattenCFGOpt::FlattenParallelAndOr(BasicBlock *BB, IRBuilder<> &Builder,
                                         Pass *P) {
  PHINode *PHI = dyn_cast<PHINode>(BB->begin());
  if (PHI)
    return false; // For simplicity, avoid cases containing PHI nodes.

  BasicBlock *LastCondBlock = NULL;
  BasicBlock *FirstCondBlock = NULL;
  BasicBlock *UnCondBlock = NULL;
  int Idx = -1;

  // Check predecessors of \param BB.
  SmallPtrSet<BasicBlock *, 16> Preds(pred_begin(BB), pred_end(BB));
  for (SmallPtrSetIterator<BasicBlock *> PI = Preds.begin(), PE = Preds.end();
       PI != PE; ++PI) {
    BasicBlock *Pred = *PI;
    BranchInst *PBI = dyn_cast<BranchInst>(Pred->getTerminator());

    // All predecessors should terminate with a branch.
    if (!PBI)
      return false;

    BasicBlock *PP = Pred->getSinglePredecessor();

    if (PBI->isUnconditional()) {
      // Case 1: Pred (BB3) is an unconditional block, it should
//.........这里部分代码省略.........

/// Check whether \param BB is the merge block of a if-region.  If yes, check
/// whether there exists an adjacent if-region upstream, the two if-regions
/// contain identical instuctions and can be legally merged.  \returns true if
/// the two if-regions are merged.
///
/// From:
/// if (a)
///   statement;
/// if (b)
///   statement;
///
/// To:
/// if (a || b)
///   statement;
///
bool FlattenCFGOpt::MergeIfRegion(BasicBlock *BB, IRBuilder<> &Builder,
                                  Pass *P) {
  BasicBlock *IfTrue2, *IfFalse2;
  Value *IfCond2 = GetIfCondition(BB, IfTrue2, IfFalse2);
  Instruction *CInst2 = dyn_cast_or_null<Instruction>(IfCond2);
  if (!CInst2)
    return false;

  BasicBlock *SecondEntryBlock = CInst2->getParent();
  if (SecondEntryBlock->hasAddressTaken())
    return false;

  BasicBlock *IfTrue1, *IfFalse1;
  Value *IfCond1 = GetIfCondition(SecondEntryBlock, IfTrue1, IfFalse1);
  Instruction *CInst1 = dyn_cast_or_null<Instruction>(IfCond1);
  if (!CInst1)
    return false;

  BasicBlock *FirstEntryBlock = CInst1->getParent();

  // Either then-path or else-path should be empty.
  if ((IfTrue1 != FirstEntryBlock) && (IfFalse1 != FirstEntryBlock))
    return false;
  if ((IfTrue2 != SecondEntryBlock) && (IfFalse2 != SecondEntryBlock))
    return false;

  TerminatorInst *PTI2 = SecondEntryBlock->getTerminator();
  Instruction *PBI2 = SecondEntryBlock->begin();

  if (!CompareIfRegionBlock(FirstEntryBlock, SecondEntryBlock, IfTrue1,
                            IfTrue2))
    return false;

  if (!CompareIfRegionBlock(FirstEntryBlock, SecondEntryBlock, IfFalse1,
                            IfFalse2))
    return false;

  // Check whether \param SecondEntryBlock has side-effect and is safe to
  // speculate.
  for (BasicBlock::iterator BI = PBI2, BE = PTI2; BI != BE; ++BI) {
    Instruction *CI = BI;
    if (isa<PHINode>(CI) || CI->mayHaveSideEffects() ||
        !isSafeToSpeculativelyExecute(CI))
      return false;
  }

  // Merge \param SecondEntryBlock into \param FirstEntryBlock.
  FirstEntryBlock->getInstList().pop_back();
  FirstEntryBlock->getInstList()
      .splice(FirstEntryBlock->end(), SecondEntryBlock->getInstList());
  BranchInst *PBI = dyn_cast<BranchInst>(FirstEntryBlock->getTerminator());
  Value *CC = PBI->getCondition();
  BasicBlock *SaveInsertBB = Builder.GetInsertBlock();
  BasicBlock::iterator SaveInsertPt = Builder.GetInsertPoint();
  Builder.SetInsertPoint(PBI);
  Value *NC = Builder.CreateOr(CInst1, CC);
  PBI->replaceUsesOfWith(CC, NC);
  Builder.SetInsertPoint(SaveInsertBB, SaveInsertPt);

  // Remove IfTrue1
  if (IfTrue1 != FirstEntryBlock) {
    IfTrue1->dropAllReferences();
    IfTrue1->eraseFromParent();
  }

  // Remove IfFalse1
  if (IfFalse1 != FirstEntryBlock) {
    IfFalse1->dropAllReferences();
    IfFalse1->eraseFromParent();
  }

  // Remove \param SecondEntryBlock
  SecondEntryBlock->dropAllReferences();
  SecondEntryBlock->eraseFromParent();
  DEBUG(dbgs() << "If conditions merged into:\n" << *FirstEntryBlock);
  return true;
}

void TripCountGenerator::generateVectorEstimatedTripCounts(Function &F){

	LoopInfoEx& li = getAnalysis<LoopInfoEx>();
	LoopNormalizerAnalysis& ln = getAnalysis<LoopNormalizerAnalysis>();

	for(LoopInfoEx::iterator lit = li.begin(); lit != li.end(); lit++){

		//Indicates if we don't have ways to determine the trip count
		bool unknownTC = false;

		Loop* loop = *lit;

		BasicBlock* header = loop->getHeader();
		BasicBlock* entryBlock = ln.entryBlocks[header];

		LoopControllersDepGraph& lcd = getAnalysis<LoopControllersDepGraph>();
		lcd.setPerspective(header);

		/*
		 * Here we are looking for the predicate that stops the loop.
		 *
		 * At this moment, we are only considering loops that are controlled by
		 * integer comparisons.
		 */
		BasicBlock* exitBlock = findLoopControllerBlock(loop);
		assert(exitBlock && "Exiting Block not found!");

		TerminatorInst* T = exitBlock->getTerminator();
		BranchInst* BI = dyn_cast<BranchInst>(T);
		ICmpInst* CI = BI ? dyn_cast<ICmpInst>(BI->getCondition()) : NULL;

		Value* Op1 = NULL;
		Value* Op2 = NULL;

		if (!CI) unknownTC = true;
		else {

			int LoopClass;
			if (isIntervalComparison(CI)) {
				LoopClass = 0;
				NumIntervalLoops++;
			} else {
				LoopClass = 1;
				NumEqualityLoops++;
			}

			Op1 = getValueAtEntryPoint(CI->getOperand(0), header);
			Op2 = getValueAtEntryPoint(CI->getOperand(1), header);


			if((!Op1) || (!Op2) ) {

				if (!LoopClass) NumUnknownConditionsIL++;
				else 			NumUnknownConditionsEL++;

				unknownTC = true;
			} else {


				if (!(Op1->getType()->isIntegerTy() && Op2->getType()->isIntegerTy())) {
					//We only handle loop conditions that compares integer variables
					NumIncompatibleOperandTypes++;
					unknownTC = true;
				}

			}

		}

		ProgressVector* V1 = NULL;
		ProgressVector* V2 = NULL;


		if (!unknownTC) {
			V1 = generateConstantProgressVector(CI->getOperand(0), header);
			V2 = generateConstantProgressVector(CI->getOperand(1), header);

			if ((!V1) || (!V2)) {

				//TODO: Increment a statistic here
				unknownTC = true;
			}

		}

		if(!unknownTC) {
			generateVectorEstimatedTripCount(header, entryBlock, Op1, Op2, V1, V2, CI);
			NumVectorEstimatedTCs++;
		}


	}

}

std::unique_ptr<FunctionOutliningInfo>
PartialInlinerImpl::computeOutliningInfo(Function *F) {
  BasicBlock *EntryBlock = &F->front();
  BranchInst *BR = dyn_cast<BranchInst>(EntryBlock->getTerminator());
  if (!BR || BR->isUnconditional())
    return std::unique_ptr<FunctionOutliningInfo>();

  // Returns true if Succ is BB's successor
  auto IsSuccessor = [](BasicBlock *Succ, BasicBlock *BB) {
    return is_contained(successors(BB), Succ);
  };

  auto SuccSize = [](BasicBlock *BB) {
    return std::distance(succ_begin(BB), succ_end(BB));
  };

  auto IsReturnBlock = [](BasicBlock *BB) {
    TerminatorInst *TI = BB->getTerminator();
    return isa<ReturnInst>(TI);
  };

  auto GetReturnBlock = [&](BasicBlock *Succ1, BasicBlock *Succ2) {
    if (IsReturnBlock(Succ1))
      return std::make_tuple(Succ1, Succ2);
    if (IsReturnBlock(Succ2))
      return std::make_tuple(Succ2, Succ1);

    return std::make_tuple<BasicBlock *, BasicBlock *>(nullptr, nullptr);
  };

  // Detect a triangular shape:
  auto GetCommonSucc = [&](BasicBlock *Succ1, BasicBlock *Succ2) {
    if (IsSuccessor(Succ1, Succ2))
      return std::make_tuple(Succ1, Succ2);
    if (IsSuccessor(Succ2, Succ1))
      return std::make_tuple(Succ2, Succ1);

    return std::make_tuple<BasicBlock *, BasicBlock *>(nullptr, nullptr);
  };

  std::unique_ptr<FunctionOutliningInfo> OutliningInfo =
      llvm::make_unique<FunctionOutliningInfo>();

  BasicBlock *CurrEntry = EntryBlock;
  bool CandidateFound = false;
  do {
    // The number of blocks to be inlined has already reached
    // the limit. When MaxNumInlineBlocks is set to 0 or 1, this
    // disables partial inlining for the function.
    if (OutliningInfo->GetNumInlinedBlocks() >= MaxNumInlineBlocks)
      break;

    if (SuccSize(CurrEntry) != 2)
      break;

    BasicBlock *Succ1 = *succ_begin(CurrEntry);
    BasicBlock *Succ2 = *(succ_begin(CurrEntry) + 1);

    BasicBlock *ReturnBlock, *NonReturnBlock;
    std::tie(ReturnBlock, NonReturnBlock) = GetReturnBlock(Succ1, Succ2);

    if (ReturnBlock) {
      OutliningInfo->Entries.push_back(CurrEntry);
      OutliningInfo->ReturnBlock = ReturnBlock;
      OutliningInfo->NonReturnBlock = NonReturnBlock;
      CandidateFound = true;
      break;
    }

    BasicBlock *CommSucc;
    BasicBlock *OtherSucc;
    std::tie(CommSucc, OtherSucc) = GetCommonSucc(Succ1, Succ2);

    if (!CommSucc)
      break;

    OutliningInfo->Entries.push_back(CurrEntry);
    CurrEntry = OtherSucc;

  } while (true);

  if (!CandidateFound)
    return std::unique_ptr<FunctionOutliningInfo>();

  // Do sanity check of the entries: threre should not
  // be any successors (not in the entry set) other than
  // {ReturnBlock, NonReturnBlock}
  assert(OutliningInfo->Entries[0] == &F->front() &&
         "Function Entry must be the first in Entries vector");
  DenseSet<BasicBlock *> Entries;
  for (BasicBlock *E : OutliningInfo->Entries)
    Entries.insert(E);

  // Returns true of BB has Predecessor which is not
  // in Entries set.
  auto HasNonEntryPred = [Entries](BasicBlock *BB) {
    for (auto Pred : predecessors(BB)) {
      if (!Entries.count(Pred))
        return true;
    }
//.........这里部分代码省略.........

/// Rotate loop LP. Return true if the loop is rotated.
bool LoopRotate::rotateLoop(Loop *L) {
  // If the loop has only one block then there is not much to rotate.
  if (L->getBlocks().size() == 1)
    return false;
  
  BasicBlock *OrigHeader = L->getHeader();
  
  BranchInst *BI = dyn_cast<BranchInst>(OrigHeader->getTerminator());
  if (BI == 0 || BI->isUnconditional())
    return false;
  
  // If the loop header is not one of the loop exiting blocks then
  // either this loop is already rotated or it is not
  // suitable for loop rotation transformations.
  if (!L->isLoopExiting(OrigHeader))
    return false;

  // Updating PHInodes in loops with multiple exits adds complexity. 
  // Keep it simple, and restrict loop rotation to loops with one exit only.
  // In future, lift this restriction and support for multiple exits if
  // required.
  SmallVector<BasicBlock*, 8> ExitBlocks;
  L->getExitBlocks(ExitBlocks);
  if (ExitBlocks.size() > 1)
    return false;

  // Check size of original header and reject loop if it is very big.
  {
    CodeMetrics Metrics;
    Metrics.analyzeBasicBlock(OrigHeader);
    if (Metrics.NumInsts > MAX_HEADER_SIZE)
      return false;
  }

  // Now, this loop is suitable for rotation.
  BasicBlock *OrigPreheader = L->getLoopPreheader();
  BasicBlock *OrigLatch = L->getLoopLatch();
  
  // If the loop could not be converted to canonical form, it must have an
  // indirectbr in it, just give up.
  if (OrigPreheader == 0 || OrigLatch == 0)
    return false;

  // Anything ScalarEvolution may know about this loop or the PHI nodes
  // in its header will soon be invalidated.
  if (ScalarEvolution *SE = getAnalysisIfAvailable<ScalarEvolution>())
    SE->forgetLoop(L);

  // Find new Loop header. NewHeader is a Header's one and only successor
  // that is inside loop.  Header's other successor is outside the
  // loop.  Otherwise loop is not suitable for rotation.
  BasicBlock *Exit = BI->getSuccessor(0);
  BasicBlock *NewHeader = BI->getSuccessor(1);
  if (L->contains(Exit))
    std::swap(Exit, NewHeader);
  assert(NewHeader && "Unable to determine new loop header");
  assert(L->contains(NewHeader) && !L->contains(Exit) && 
         "Unable to determine loop header and exit blocks");
  
  // This code assumes that the new header has exactly one predecessor.
  // Remove any single-entry PHI nodes in it.
  assert(NewHeader->getSinglePredecessor() &&
         "New header doesn't have one pred!");
  FoldSingleEntryPHINodes(NewHeader);

  // Begin by walking OrigHeader and populating ValueMap with an entry for
  // each Instruction.
  BasicBlock::iterator I = OrigHeader->begin(), E = OrigHeader->end();
  ValueToValueMapTy ValueMap;

  // For PHI nodes, the value available in OldPreHeader is just the
  // incoming value from OldPreHeader.
  for (; PHINode *PN = dyn_cast<PHINode>(I); ++I)
    ValueMap[PN] = PN->getIncomingValue(PN->getBasicBlockIndex(OrigPreheader));

  // For the rest of the instructions, either hoist to the OrigPreheader if
  // possible or create a clone in the OldPreHeader if not.
  TerminatorInst *LoopEntryBranch = OrigPreheader->getTerminator();
  while (I != E) {
    Instruction *Inst = I++;
    
    // If the instruction's operands are invariant and it doesn't read or write
    // memory, then it is safe to hoist.  Doing this doesn't change the order of
    // execution in the preheader, but does prevent the instruction from
    // executing in each iteration of the loop.  This means it is safe to hoist
    // something that might trap, but isn't safe to hoist something that reads
    // memory (without proving that the loop doesn't write).
    if (L->hasLoopInvariantOperands(Inst) &&
        !Inst->mayReadFromMemory() && !Inst->mayWriteToMemory() &&
        !isa<TerminatorInst>(Inst) && !isa<DbgInfoIntrinsic>(Inst)) {
      Inst->moveBefore(LoopEntryBranch);
      continue;
    }
    
    // Otherwise, create a duplicate of the instruction.
    Instruction *C = Inst->clone();
    
    // Eagerly remap the operands of the instruction.
    RemapInstruction(C, ValueMap,
//.........这里部分代码省略.........

bool BranchProbabilityInfo::calcZeroHeuristics(BasicBlock *BB) {
  BranchInst * BI = dyn_cast<BranchInst>(BB->getTerminator());
  if (!BI || !BI->isConditional())
    return false;

  Value *Cond = BI->getCondition();
  ICmpInst *CI = dyn_cast<ICmpInst>(Cond);
  if (!CI)
    return false;

  Value *RHS = CI->getOperand(1);
  ConstantInt *CV = dyn_cast<ConstantInt>(RHS);
  if (!CV)
    return false;

  bool isProb;
  if (CV->isZero()) {
    switch (CI->getPredicate()) {
    case CmpInst::ICMP_EQ:
      // X == 0   ->  Unlikely
      isProb = false;
      break;
    case CmpInst::ICMP_NE:
      // X != 0   ->  Likely
      isProb = true;
      break;
    case CmpInst::ICMP_SLT:
      // X < 0   ->  Unlikely
      isProb = false;
      break;
    case CmpInst::ICMP_SGT:
      // X > 0   ->  Likely
      isProb = true;
      break;
    default:
      return false;
    }
  } else if (CV->isOne() && CI->getPredicate() == CmpInst::ICMP_SLT) {
    // InstCombine canonicalizes X <= 0 into X < 1.
    // X <= 0   ->  Unlikely
    isProb = false;
  } else if (CV->isAllOnesValue()) {
    switch (CI->getPredicate()) {
    case CmpInst::ICMP_EQ:
      // X == -1  ->  Unlikely
      isProb = false;
      break;
    case CmpInst::ICMP_NE:
      // X != -1  ->  Likely
      isProb = true;
      break;
    case CmpInst::ICMP_SGT:
      // InstCombine canonicalizes X >= 0 into X > -1.
      // X >= 0   ->  Likely
      isProb = true;
      break;
    default:
      return false;
    }
  } else {
    return false;
  }

  unsigned TakenIdx = 0, NonTakenIdx = 1;

  if (!isProb)
    std::swap(TakenIdx, NonTakenIdx);

  setEdgeWeight(BB, TakenIdx, ZH_TAKEN_WEIGHT);
  setEdgeWeight(BB, NonTakenIdx, ZH_NONTAKEN_WEIGHT);

  return true;
}

bool BranchProbabilityInfo::calcZeroHeuristics(BasicBlock *BB) {
    BranchInst * BI = dyn_cast<BranchInst>(BB->getTerminator());
    if (!BI || !BI->isConditional())
        return false;

    Value *Cond = BI->getCondition();
    ICmpInst *CI = dyn_cast<ICmpInst>(Cond);
    if (!CI)
        return false;

    Value *RHS = CI->getOperand(1);
    ConstantInt *CV = dyn_cast<ConstantInt>(RHS);
    if (!CV)
        return false;

    // If the LHS is the result of AND'ing a value with a single bit bitmask,
    // we don't have information about probabilities.
    if (Instruction *LHS = dyn_cast<Instruction>(CI->getOperand(0)))
        if (LHS->getOpcode() == Instruction::And)
            if (ConstantInt *AndRHS = dyn_cast<ConstantInt>(LHS->getOperand(1)))
                if (AndRHS->getUniqueInteger().isPowerOf2())
                    return false;

    bool isProb;
    if (CV->isZero()) {
        switch (CI->getPredicate()) {
        case CmpInst::ICMP_EQ:
            // X == 0   ->  Unlikely
            isProb = false;
            break;
        case CmpInst::ICMP_NE:
            // X != 0   ->  Likely
            isProb = true;
            break;
        case CmpInst::ICMP_SLT:
            // X < 0   ->  Unlikely
            isProb = false;
            break;
        case CmpInst::ICMP_SGT:
            // X > 0   ->  Likely
            isProb = true;
            break;
        default:
            return false;
        }
    } else if (CV->isOne() && CI->getPredicate() == CmpInst::ICMP_SLT) {
        // InstCombine canonicalizes X <= 0 into X < 1.
        // X <= 0   ->  Unlikely
        isProb = false;
    } else if (CV->isAllOnesValue()) {
        switch (CI->getPredicate()) {
        case CmpInst::ICMP_EQ:
            // X == -1  ->  Unlikely
            isProb = false;
            break;
        case CmpInst::ICMP_NE:
            // X != -1  ->  Likely
            isProb = true;
            break;
        case CmpInst::ICMP_SGT:
            // InstCombine canonicalizes X >= 0 into X > -1.
            // X >= 0   ->  Likely
            isProb = true;
            break;
        default:
            return false;
        }
    } else {
        return false;
    }

    unsigned TakenIdx = 0, NonTakenIdx = 1;

    if (!isProb)
        std::swap(TakenIdx, NonTakenIdx);

    setEdgeWeight(BB, TakenIdx, ZH_TAKEN_WEIGHT);
    setEdgeWeight(BB, NonTakenIdx, ZH_NONTAKEN_WEIGHT);

    return true;
}

//.........这里部分代码省略.........
    if (BCI)
      V = BCI->getOperand(0);

    PN = dyn_cast<PHINode>(V);
    if (!PN)
      return false;
  }

  BasicBlock *BB = RI->getParent();
  if (PN && PN->getParent() != BB)
    return false;

  // It's not safe to eliminate the sign / zero extension of the return value.
  // See llvm::isInTailCallPosition().
  const Function *F = BB->getParent();
  Attributes CallerRetAttr = F->getAttributes().getRetAttributes();
  if (CallerRetAttr.hasAttribute(Attributes::ZExt) ||
      CallerRetAttr.hasAttribute(Attributes::SExt))
    return false;

  // Make sure there are no instructions between the PHI and return, or that the
  // return is the first instruction in the block.
  if (PN) {
    BasicBlock::iterator BI = BB->begin();
    do { ++BI; } while (isa<DbgInfoIntrinsic>(BI));
    if (&*BI == BCI)
      // Also skip over the bitcast.
      ++BI;
    if (&*BI != RI)
      return false;
  } else {
    BasicBlock::iterator BI = BB->begin();
    while (isa<DbgInfoIntrinsic>(BI)) ++BI;
    if (&*BI != RI)
      return false;
  }

  /// Only dup the ReturnInst if the CallInst is likely to be emitted as a tail
  /// call.
  SmallVector<CallInst*, 4> TailCalls;
  if (PN) {
    for (unsigned I = 0, E = PN->getNumIncomingValues(); I != E; ++I) {
      CallInst *CI = dyn_cast<CallInst>(PN->getIncomingValue(I));
      // Make sure the phi value is indeed produced by the tail call.
      if (CI && CI->hasOneUse() && CI->getParent() == PN->getIncomingBlock(I) &&
          TLI->mayBeEmittedAsTailCall(CI))
        TailCalls.push_back(CI);
    }
  } else {
    SmallPtrSet<BasicBlock*, 4> VisitedBBs;
    for (pred_iterator PI = pred_begin(BB), PE = pred_end(BB); PI != PE; ++PI) {
      if (!VisitedBBs.insert(*PI))
        continue;

      BasicBlock::InstListType &InstList = (*PI)->getInstList();
      BasicBlock::InstListType::reverse_iterator RI = InstList.rbegin();
      BasicBlock::InstListType::reverse_iterator RE = InstList.rend();
      do { ++RI; } while (RI != RE && isa<DbgInfoIntrinsic>(&*RI));
      if (RI == RE)
        continue;

      CallInst *CI = dyn_cast<CallInst>(&*RI);
      if (CI && CI->use_empty() && TLI->mayBeEmittedAsTailCall(CI))
        TailCalls.push_back(CI);
    }
  }

  bool Changed = false;
  for (unsigned i = 0, e = TailCalls.size(); i != e; ++i) {
    CallInst *CI = TailCalls[i];
    CallSite CS(CI);

    // Conservatively require the attributes of the call to match those of the
    // return. Ignore noalias because it doesn't affect the call sequence.
    Attributes CalleeRetAttr = CS.getAttributes().getRetAttributes();
    if (AttrBuilder(CalleeRetAttr).
          removeAttribute(Attributes::NoAlias) !=
        AttrBuilder(CallerRetAttr).
          removeAttribute(Attributes::NoAlias))
      continue;

    // Make sure the call instruction is followed by an unconditional branch to
    // the return block.
    BasicBlock *CallBB = CI->getParent();
    BranchInst *BI = dyn_cast<BranchInst>(CallBB->getTerminator());
    if (!BI || !BI->isUnconditional() || BI->getSuccessor(0) != BB)
      continue;

    // Duplicate the return into CallBB.
    (void)FoldReturnIntoUncondBranch(RI, BB, CallBB);
    ModifiedDT = Changed = true;
    ++NumRetsDup;
  }

  // If we eliminated all predecessors of the block, delete the block now.
  if (Changed && !BB->hasAddressTaken() && pred_begin(BB) == pred_end(BB))
    BB->eraseFromParent();

  return Changed;
}