[LV] Unconditionally branch from middle to scalar preheader if the scalar loop must execute

[reames]

Summary

If we know that the scalar epilogue is required to run, modify the CFG to end the middle block with an unconditional branch to scalar preheader. This is instead of a conditional branch to either the preheader or the exit block.

The motivation to do this is to support multiple exit blocks. Specifically, the current structure forces us to identify immediate dominators and *which* exit block to branch from in the middle terminator. For the multiple exit case - where we know require scalar will hold - these questions are ill formed.

This is the last change needed to support multiple exit loops, but since the diffs are already large enough, I'm going to land this, and then enable separately. You can think of this as being NFCIish prep work, but the changes are a bit too involved for me to feel comfortable tagging the review that way.

[reames]

Update ascii art to reflect change and clarify a confusing bit in the original.

[reames]

rebase over a landed set of tests

[fhahn]

This makes sense, but the fact that we need to scatter tests throughout various parts of the code seems unfortunate. I'm planning to take a closer look over the next few days to see if I can provide any ideas on how this could be improved.

[reames]

Florian, have you had a chance to give thought to alternatives? Given it's been two weeks, unless you have an actionable suggestion, I'd like to move forward with the current patch. I'll emphasize that I'm happy to iterate on the design here, either during review, or after submission if you think of something cleaner.

[fhahn]

In D94892#2534445, @reames wrote:

Florian, have you had a chance to give thought to alternatives? Given it's been two weeks, unless you have an actionable suggestion, I'd like to move forward with the current patch. I'll emphasize that I'm happy to iterate on the design here, either during review, or after submission if you think of something cleaner.

Unfortunately I did not really have time to think much about better alternatives so far (and it's unlikely I'll get time until end of next week). But I agree, it's not a really big deal and it works well enough for now. Let's tweak it post-commit, should a better alternative presents itself. So for now I just have a few small suggestions, mostly wording related.

LGTM, thanks!

llvm/lib/Transforms/Vectorize/LoopVectorize.cpp

3109

This applies for all loops that require scalar epilogues, not just ones with multiple exits, right? If so, it might be better word it in terms of requiring scalar epiloge, rather than multi exits. (perhaps something like An edge from the middle block to the exit block is only added if the scalar epilogue may not be executed. Thus only update the immediate dominators if the scalar epilogue is not required.)

3357

nit: set up a conditional branch from the middle block to the scalar loop pre-header?

3365

Do we need this assert? There's a similar assert just a few lines above. If it's not needed, getting rid of the lambda would make things a bit easier to read IMO (perhaps it could be just BranchInst *BrInst = Cost->requiresScalarEpilogue() ? BranchInst::Create(LoopScalarPreHeader) : BranchInst::Create(LoopExitBlock, LoopScalarPreHeader, Builder.getTrue()))

3483

nit: if we require *a* scalar epilogue .... as we unconditionally branch to *the* scalar preheader?

3626

nit: unrelated whitespace change

[Ayal]

llvm/lib/Transforms/Vectorize/LoopVectorize.cpp

5869

Not sure if this may help with the reported failures, but if vectorizing a loop having multiple exits is not (yet) intended to work for epilog vectorization, perhaps return false here if OrigLoop has multiple exit blocks.
(Otherwise the CFG of https://llvm.org/docs/Vectorizers.html#epilogue-vectorization may be updated).

Another thought to try and reduce the effect of the patch temporarily, is to check
if (!LoopExitBlock)
instead of
if (!Cost->requiresScalarEpilogue())
leaving the suboptimal but currently working code for single exit loops that require scalar epilogue.

[reames]

llvm/lib/Transforms/Vectorize/LoopVectorize.cpp

5869

Not sure if this may help with the reported failures, but if vectorizing a loop having multiple exits is not (yet) intended to work for epilog vectorization, perhaps return false here if OrigLoop has multiple exit blocks.
(Otherwise the CFG of https://llvm.org/docs/Vectorizers.html#epilogue-vectorization may be updated).

This seems to be a comment which applies to the future patch which enables multiple exit vectorization, not this one. Unless I'm missing something?

Also, I would really expect a flag called *requires* scalar epilogue to override the epilogue vectorization setting, but I haven't stared at the code enough to know if that's really true.

llvm/lib/Transforms/Utils/LoopVersioning.cpp

 LoopVersioning::LoopVersioning(const LoopAccessInfo &LAI,
                                ArrayRef<RuntimePointerCheck> Checks, Loop *L,
                                LoopInfo *LI, DominatorTree *DT,
                                ScalarEvolution *SE)
     : VersionedLoop(L), NonVersionedLoop(nullptr),
       AliasChecks(Checks.begin(), Checks.end()),
       Preds(LAI.getPSE().getUnionPredicate()), LAI(LAI), LI(LI), DT(DT),
       SE(SE) {
-  assert(L->getUniqueExitBlock() && "No single exit block");
 }
 
 void LoopVersioning::versionLoop(
     const SmallVectorImpl<Instruction *> &DefsUsedOutside) {
+  assert(VersionedLoop->getUniqueExitBlock() && "No single exit block");
   assert(VersionedLoop->isLoopSimplifyForm() &&
          "Loop is not in loop-simplify form");
 
   Instruction *FirstCheckInst;
   Instruction *MemRuntimeCheck;
   Value *SCEVRuntimeCheck;
   Value *RuntimeCheck = nullptr;
 

llvm/lib/Transforms/Vectorize/LoopVectorize.cpp

protected:
   BasicBlock *LoopVectorPreHeader;
 
   /// The scalar-loop preheader.
   BasicBlock *LoopScalarPreHeader;
 
   /// Middle Block between the vector and the scalar.
   BasicBlock *LoopMiddleBlock;
 
-  /// The (unique) ExitBlock of the scalar loop.  Note that
+  /// The unique ExitBlock of the scalar loop if one exists.  Note that
   /// there can be multiple exiting edges reaching this block.
   BasicBlock *LoopExitBlock;
 
   /// The vector loop body.
   BasicBlock *LoopVectorBody;
 
   /// The scalar loop body.
   BasicBlock *LoopScalarBody;
void InnerLoopVectorizer::emitMinimumIterationCountCheck(Loop *L,
   LoopVectorPreHeader =
       SplitBlock(TCCheckBlock, TCCheckBlock->getTerminator(), DT, LI, nullptr,
                  "vector.ph");
 
   assert(DT->properlyDominates(DT->getNode(TCCheckBlock),
                                DT->getNode(Bypass)->getIDom()) &&
          "TC check is expected to dominate Bypass");
 
-  // Update dominator for Bypass & LoopExit.
+  // Update dominator for Bypass & LoopExit (if needed).
   DT->changeImmediateDominator(Bypass, TCCheckBlock);
+  if (!Cost->requiresScalarEpilogue())
+    // For loops with multiple exits, there's no edge from the middle block
+    // to exit blocks (as the epilogue must run) and thus no need to update
+    // the immediate dominator of the exit blocks.
     DT->changeImmediateDominator(LoopExitBlock, TCCheckBlock);
 
   ReplaceInstWithInst(
       TCCheckBlock->getTerminator(),
       BranchInst::Create(Bypass, LoopVectorPreHeader, CheckMinIters));
   LoopBypassBlocks.push_back(TCCheckBlock);
 }
 
 void InnerLoopVectorizer::emitSCEVChecks(Loop *L, BasicBlock *Bypass) {
void InnerLoopVectorizer::emitSCEVChecks(Loop *L, BasicBlock *Bypass) {
   // Create new preheader for vector loop.
   LoopVectorPreHeader =
       SplitBlock(SCEVCheckBlock, SCEVCheckBlock->getTerminator(), DT, LI,
                  nullptr, "vector.ph");
 
   // Update dominator only if this is first RT check.
   if (LoopBypassBlocks.empty()) {
     DT->changeImmediateDominator(Bypass, SCEVCheckBlock);
+    if (!Cost->requiresScalarEpilogue())
+      // For loops with multiple exits, there's no edge from the middle block
+      // to exit blocks (as the epilogue must run) and thus no need to update
+      // the immediate dominator of the exit blocks.
       DT->changeImmediateDominator(LoopExitBlock, SCEVCheckBlock);
   }
 
   ReplaceInstWithInst(
       SCEVCheckBlock->getTerminator(),
       BranchInst::Create(Bypass, LoopVectorPreHeader, SCEVCheck));
   LoopBypassBlocks.push_back(SCEVCheckBlock);
   AddedSafetyChecks = true;
 }
auto *CondBranch = cast<BranchInst>(
       Builder.CreateCondBr(Builder.getTrue(), Bypass, LoopVectorPreHeader));
   ReplaceInstWithInst(MemCheckBlock->getTerminator(), CondBranch);
   LoopBypassBlocks.push_back(MemCheckBlock);
   AddedSafetyChecks = true;
 
   // Update dominator only if this is first RT check.
   if (LoopBypassBlocks.empty()) {
     DT->changeImmediateDominator(Bypass, MemCheckBlock);
+    if (!Cost->requiresScalarEpilogue())
+      // For loops with multiple exits, there's no edge from the middle block
+      // to exit blocks (as the epilogue must run) and thus no need to update
+      // the immediate dominator of the exit blocks.
       DT->changeImmediateDominator(LoopExitBlock, MemCheckBlock);
   }
 
   Instruction *FirstCheckInst;
   Instruction *MemRuntimeCheck;
   std::tie(FirstCheckInst, MemRuntimeCheck) =
       addRuntimeChecks(MemCheckBlock->getTerminator(), OrigLoop,
                        RtPtrChecking.getChecks(), RtPtrChecking.getSE());
   assert(MemRuntimeCheck && "no RT checks generated although RtPtrChecking "
case InductionDescriptor::IK_NoInduction:
     return nullptr;
   }
   llvm_unreachable("invalid enum");
 }
 
 Loop *InnerLoopVectorizer::createVectorLoopSkeleton(StringRef Prefix) {
   LoopScalarBody = OrigLoop->getHeader();
   LoopVectorPreHeader = OrigLoop->getLoopPreheader();
-  LoopExitBlock = OrigLoop->getUniqueExitBlock();
-  assert(LoopExitBlock && "Must have an exit block");
   assert(LoopVectorPreHeader && "Invalid loop structure");
+  LoopExitBlock = OrigLoop->getUniqueExitBlock(); // may be nullptr
+  assert((LoopExitBlock || Cost->requiresScalarEpilogue()) &&
+         "multiple exit loop without required epilogue?");
 
   LoopMiddleBlock =
       SplitBlock(LoopVectorPreHeader, LoopVectorPreHeader->getTerminator(), DT,
                  LI, nullptr, Twine(Prefix) + "middle.block");
   LoopScalarPreHeader =
       SplitBlock(LoopMiddleBlock, LoopMiddleBlock->getTerminator(), DT, LI,
                  nullptr, Twine(Prefix) + "scalar.ph");
 
-  // Set up branch from middle block to the exit and scalar preheader blocks.
-  // completeLoopSkeleton will update the condition to use an iteration check,
-  // if required to decide whether to execute the remainder.
-  BranchInst *BrInst =
-      BranchInst::Create(LoopExitBlock, LoopScalarPreHeader, Builder.getTrue());
   auto *ScalarLatchTerm = OrigLoop->getLoopLatch()->getTerminator();
+
+  // Set up the middle block terminator.  Two cases:
+  // 1) If we know that we must execute the scalar epilogue, emit an
+  //    unconditional branch.
+  // 2) Otherwise, we must have a single unique exit block (due to how we
+  //    implement the multiple exit case).  In this case, set up conditonal
+  //    branch from middle block to loop scalar preheader, and the exit block.
+  //    completeLoopSkeleton will update the condition to use an iteration
+  //   check, if required to decide whether to execute the remainder.
+  BranchInst *BrInst = [&]() {
+    if (Cost->requiresScalarEpilogue()) {
+       return BranchInst::Create(LoopScalarPreHeader);
+    }
+    assert(LoopExitBlock && "Must have an exit block");
+    return BranchInst::Create(LoopExitBlock, LoopScalarPreHeader,
+                              Builder.getTrue());
+  }();
   BrInst->setDebugLoc(ScalarLatchTerm->getDebugLoc());
   ReplaceInstWithInst(LoopMiddleBlock->getTerminator(), BrInst);
 
   // We intentionally don't let SplitBlock to update LoopInfo since
   // LoopVectorBody should belong to another loop than LoopVectorPreHeader.
   // LoopVectorBody is explicitly added to the correct place few lines later.
   LoopVectorBody =
       SplitBlock(LoopVectorPreHeader, LoopVectorPreHeader->getTerminator(), DT,
                  nullptr, nullptr, Twine(Prefix) + "vector.body");
 
   // Update dominator for loop exit.
+  if (!Cost->requiresScalarEpilogue())
+    // For loops with multiple exits, there's no edge from the middle block
+    // to exit blocks (as the epilogue must run) and thus no need to update
+    // the immediate dominator of the exit blocks.
     DT->changeImmediateDominator(LoopExitBlock, LoopMiddleBlock);
 
   // Create and register the new vector loop.
   Loop *Lp = LI->AllocateLoop();
   Loop *ParentLoop = OrigLoop->getParentLoop();
 
   // Insert the new loop into the loop nest and register the new basic blocks
   // before calling any utilities such as SCEV that require valid LoopInfo.
   if (ParentLoop) {
BasicBlock *InnerLoopVectorizer::completeLoopSkeleton(Loop *L,
 
   // The trip counts should be cached by now.
   Value *Count = getOrCreateTripCount(L);
   Value *VectorTripCount = getOrCreateVectorTripCount(L);
 
   auto *ScalarLatchTerm = OrigLoop->getLoopLatch()->getTerminator();
 
   // Add a check in the middle block to see if we have completed
-  // all of the iterations in the first vector loop.
-  // If (N - N%VF) == N, then we *don't* need to run the remainder.
-  // If tail is to be folded, we know we don't need to run the remainder.
-  if (!Cost->foldTailByMasking()) {
+  // all of the iterations in the first vector loop.  Three cases:
+  // 1) If we require scalar epilogue, there is no conditional branch as
+  //    we unconditionally branch to scalar preheader.  Do nothing.
+  // 2) If (N - N%VF) == N, then we *don't* need to run the remainder.
+  //    Thus if tail is to be folded, we know we don't need to run the
+  //    remainder and we can use the previous value for the condition (true).
+  // 3) Otherwise, construct a runtime check.
+  if (!Cost->requiresScalarEpilogue() && !Cost->foldTailByMasking()) {
     Instruction *CmpN = CmpInst::Create(Instruction::ICmp, CmpInst::ICMP_EQ,
                                         Count, VectorTripCount, "cmp.n",
                                         LoopMiddleBlock->getTerminator());
 
     // Here we use the same DebugLoc as the scalar loop latch terminator instead
     // of the corresponding compare because they may have ended up with
     // different line numbers and we want to avoid awkward line stepping while
     // debugging. Eg. if the compare has got a line number inside the loop.

 // Fix up external users of the induction variable. At this point, we are
 // in LCSSA form, with all external PHIs that use the IV having one input value,
 // coming from the remainder loop. We need those PHIs to also have a correct
 // value for the IV when arriving directly from the middle block.
 void InnerLoopVectorizer::fixupIVUsers(PHINode *OrigPhi,
                                        const InductionDescriptor &II,
                                        Value *CountRoundDown, Value *EndValue,
                                        BasicBlock *MiddleBlock) {
+
   // There are two kinds of external IV usages - those that use the value
   // computed in the last iteration (the PHI) and those that use the penultimate
   // value (the value that feeds into the phi from the loop latch).
   // We allow both, but they, obviously, have different values.
 
   assert(OrigLoop->getUniqueExitBlock() && "Expected a single exit block");
 
   DenseMap<Value *, Value *> MissingVals;
void InnerLoopVectorizer::fixVectorizedLoop() {
   // vector form. Now we need to fix the recurrences in the loop. These PHI
   // nodes are currently empty because we did not want to introduce cycles.
   // This is the second stage of vectorizing recurrences.
   fixCrossIterationPHIs();
 
   // Forget the original basic block.
   PSE.getSE()->forgetLoop(OrigLoop);
 
+  // If we inserted an edge from the middle block to the unique exit block,
+  // update uses outside the loop (phis) to account for the newly inserted
+  // edge.
+  if (!Cost->requiresScalarEpilogue()) {
     // Fix-up external users of the induction variables.
     for (auto &Entry : Legal->getInductionVars())
       fixupIVUsers(Entry.first, Entry.second,
                    getOrCreateVectorTripCount(LI->getLoopFor(LoopVectorBody)),
                    IVEndValues[Entry.first], LoopMiddleBlock);
 
     fixLCSSAPHIs();
+  }
   for (Instruction *PI : PredicatedInstructions)
     sinkScalarOperands(&*PI);
 
   // Remove redundant induction instructions.
   cse(LoopVectorBody);
 
   // Set/update profile weights for the vector and remainder loops as original
   // loop iterations are now distributed among them. Note that original loop
void InnerLoopVectorizer::fixFirstOrderRecurrence(PHINode *Phi) {
 
   // Finally, fix users of the recurrence outside the loop. The users will need
   // either the last value of the scalar recurrence or the last value of the
   // vector recurrence we extracted in the middle block. Since the loop is in
   // LCSSA form, we just need to find all the phi nodes for the original scalar
   // recurrence in the exit block, and then add an edge for the middle block.
   // Note that LCSSA does not imply single entry when the original scalar loop
   // had multiple exiting edges (as we always run the last iteration in the
-  // scalar epilogue); in that case, the exiting path through middle will be
-  // dynamically dead and the value picked for the phi doesn't matter.
+  // scalar epilogue); in that case, there is no edge from middle to exit and
+  // and thus no phis which needed updated.
+  if (!Cost->requiresScalarEpilogue())
     for (PHINode &LCSSAPhi : LoopExitBlock->phis())
       if (any_of(LCSSAPhi.incoming_values(),
                  [Phi](Value *V) { return V == Phi; }))
         LCSSAPhi.addIncoming(ExtractForPhiUsedOutsideLoop, LoopMiddleBlock);
 }
 
 void InnerLoopVectorizer::fixReduction(PHINode *Phi) {
   // Get it's reduction variable descriptor.
   assert(Legal->isReductionVariable(Phi) &&
          "Unable to find the reduction variable");
   RecurrenceDescriptor RdxDesc = Legal->getReductionVars()[Phi];
 
void InnerLoopVectorizer::fixReduction(PHINode *Phi) {
   BCBlockPhi->addIncoming(ReducedPartRdx, LoopMiddleBlock);
 
   // Now, we need to fix the users of the reduction variable
   // inside and outside of the scalar remainder loop.
 
   // We know that the loop is in LCSSA form. We need to update the PHI nodes
   // in the exit blocks.  See comment on analogous loop in
   // fixFirstOrderRecurrence for a more complete explaination of the logic.
+  if (!Cost->requiresScalarEpilogue())
     for (PHINode &LCSSAPhi : LoopExitBlock->phis())
       if (any_of(LCSSAPhi.incoming_values(),
                  [LoopExitInst](Value *V) { return V == LoopExitInst; }))
         LCSSAPhi.addIncoming(ReducedPartRdx, LoopMiddleBlock);
 
   // Fix the scalar loop reduction variable with the incoming reduction sum
   // from the vector body and from the backedge value.
   int IncomingEdgeBlockIdx =
     Phi->getBasicBlockIndex(OrigLoop->getLoopLatch());
   assert(IncomingEdgeBlockIdx >= 0 && "Invalid block index");
   // Pick the other block.
   int SelfEdgeBlockIdx = (IncomingEdgeBlockIdx ? 0 : 1);
if (LCSSAPhi.getBasicBlockIndex(LoopMiddleBlock) != -1)
       // code above, leave them alone.
       continue;
 
     auto *IncomingValue = LCSSAPhi.getIncomingValue(0);
     // Non-instruction incoming values will have only one value.
     unsigned LastLane = 0;
     if (isa<Instruction>(IncomingValue))
       LastLane = Cost->isUniformAfterVectorization(
-                     cast<Instruction>(IncomingValue), VF)
-                     ? 0
-                     : VF.getKnownMinValue() - 1;
     assert((!VF.isScalable() || LastLane == 0) &&
            "scalable vectors dont support non-uniform scalars yet");
     // Can be a loop invariant incoming value or the last scalar value to be
     // extracted from the vectorized loop.
     Builder.SetInsertPoint(LoopMiddleBlock->getTerminator());
     Value *lastIncomingValue =
       getOrCreateScalarValue(IncomingValue, { UF - 1, LastLane });
     LCSSAPhi.addIncoming(lastIncomingValue, LoopMiddleBlock);
LLVM_DEBUG(if (ForceVectorization && Width > 1 && Cost >= ScalarCost) dbgs()
              << "LV: Vectorization seems to be not beneficial, "
              << "but was forced by a user.\n");
   LLVM_DEBUG(dbgs() << "LV: Selecting VF: " << Width << ".\n");
   VectorizationFactor Factor = {ElementCount::getFixed(Width),
                                 (unsigned)(Width * Cost)};
   return Factor;
 }
 
 bool LoopVectorizationCostModel::isCandidateForEpilogueVectorization(
     const Loop &L, ElementCount VF) const {
   // Cross iteration phis such as reductions need special handling and are
   // currently unsupported.
   if (any_of(L.getHeader()->phis(), [&](PHINode &Phi) {
         return Legal->isFirstOrderRecurrence(&Phi) ||
                Legal->isReductionVariable(&Phi);
       }))
     return false;
BasicBlock *EpilogueVectorizerMainLoop::emitMinimumIterationCountCheck(
 
   if (ForEpilogue) {
     assert(DT->properlyDominates(DT->getNode(TCCheckBlock),
                                  DT->getNode(Bypass)->getIDom()) &&
            "TC check is expected to dominate Bypass");
 
     // Update dominator for Bypass & LoopExit.
     DT->changeImmediateDominator(Bypass, TCCheckBlock);
+    if (!Cost->requiresScalarEpilogue())
+      // For loops with multiple exits, there's no edge from the middle block
+      // to exit blocks (as the epilogue must run) and thus no need to update
+      // the immediate dominator of the exit blocks.
       DT->changeImmediateDominator(LoopExitBlock, TCCheckBlock);
 
     LoopBypassBlocks.push_back(TCCheckBlock);
 
     // Save the trip count so we don't have to regenerate it in the
     // vec.epilog.iter.check. This is safe to do because the trip count
     // generated here dominates the vector epilog iter check.
     EPI.TripCount = Count;
   }
EPI.MemSafetyCheck->getTerminator()->replaceUsesOfWith(
         VecEpilogueIterationCountCheck, LoopScalarPreHeader);
 
   DT->changeImmediateDominator(
       VecEpilogueIterationCountCheck,
       VecEpilogueIterationCountCheck->getSinglePredecessor());
 
   DT->changeImmediateDominator(LoopScalarPreHeader,
                                EPI.EpilogueIterationCountCheck);
-  DT->changeImmediateDominator(LoopExitBlock, EPI.EpilogueIterationCountCheck);
+  if (!Cost->requiresScalarEpilogue())
+    // For loops with multiple exits, there's no edge from the middle block
+    // to exit blocks (as the epilogue must run) and thus no need to update
+    // the immediate dominator of the exit blocks.
+    DT->changeImmediateDominator(LoopExitBlock,
+                                 EPI.EpilogueIterationCountCheck);
 
   // Keep track of bypass blocks, as they feed start values to the induction
   // phis in the scalar loop preheader.
   if (EPI.SCEVSafetyCheck)
     LoopBypassBlocks.push_back(EPI.SCEVSafetyCheck);
   if (EPI.MemSafetyCheck)
     LoopBypassBlocks.push_back(EPI.MemSafetyCheck);
   LoopBypassBlocks.push_back(EPI.EpilogueIterationCountCheck);

llvm/test/Transforms/LoopVectorize/first-order-recurrence-complex.ll

 ; CHECK-NEXT:    [[TMP13:%.*]] = shufflevector <4 x i16> [[VECTOR_RECUR]], <4 x i16> [[WIDE_LOAD]], <4 x i32> <i32 3, i32 4, i32 5, i32 6>
 ; CHECK-NEXT:    [[TMP14:%.*]] = bitcast i16* [[TMP11]] to <4 x i16>*
 ; CHECK-NEXT:    store <4 x i16> [[TMP13]], <4 x i16>* [[TMP14]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i32 [[INDEX]], 4
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <4 x i32> [[VEC_IND]], <i32 4, i32 4, i32 4, i32 4>
 ; CHECK-NEXT:    [[TMP15:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP6:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i32 [[TMP2]], [[N_VEC]]
 ; CHECK-NEXT:    [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3
 ; CHECK-NEXT:    [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 2
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[IF_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ]
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ]
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[SCALAR_RECUR:%.*]] = phi i16 [ [[SCALAR_RECUR_INIT]], [[SCALAR_PH]] ], [ [[REC_NEXT:%.*]], [[FOR_BODY]] ]
 ; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
 ; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P]], i64 [[IPROM]]
 ; CHECK-NEXT:    [[REC_NEXT]] = load i16, i16* [[B]], align 2
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    store i16 [[SCALAR_RECUR]], i16* [[B]], align 4
 ; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
 ; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END]], [[LOOP7:!llvm.loop !.*]]
 ; CHECK:       if.end:
-; CHECK-NEXT:    [[REC_LCSSA:%.*]] = phi i16 [ [[SCALAR_RECUR]], [[FOR_BODY]] ], [ [[SCALAR_RECUR]], [[FOR_COND]] ], [ [[VECTOR_RECUR_EXTRACT_FOR_PHI]], [[MIDDLE_BLOCK]] ]
+; CHECK-NEXT:    [[REC_LCSSA:%.*]] = phi i16 [ [[SCALAR_RECUR]], [[FOR_BODY]] ], [ [[SCALAR_RECUR]], [[FOR_COND]] ]
 ; CHECK-NEXT:    ret i16 [[REC_LCSSA]]
 ;
 entry:
   br label %for.cond
 
 for.cond:
   %i = phi i32 [ 0, %entry ], [ %inc, %for.body ]
   %rec = phi i16 [0, %entry], [ %rec.next, %for.body ]

 ; CHECK-NEXT:    [[TMP13:%.*]] = shufflevector <4 x i16> [[VECTOR_RECUR]], <4 x i16> [[WIDE_LOAD]], <4 x i32> <i32 3, i32 4, i32 5, i32 6>
 ; CHECK-NEXT:    [[TMP14:%.*]] = bitcast i16* [[TMP11]] to <4 x i16>*
 ; CHECK-NEXT:    store <4 x i16> [[TMP13]], <4 x i16>* [[TMP14]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i32 [[INDEX]], 4
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <4 x i32> [[VEC_IND]], <i32 4, i32 4, i32 4, i32 4>
 ; CHECK-NEXT:    [[TMP15:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP8:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i32 [[TMP2]], [[N_VEC]]
 ; CHECK-NEXT:    [[VECTOR_RECUR_EXTRACT:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 3
 ; CHECK-NEXT:    [[VECTOR_RECUR_EXTRACT_FOR_PHI:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 2
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[IF_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[SCALAR_RECUR_INIT:%.*]] = phi i16 [ 0, [[ENTRY:%.*]] ], [ [[VECTOR_RECUR_EXTRACT]], [[MIDDLE_BLOCK]] ]
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ]
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[SCALAR_RECUR:%.*]] = phi i16 [ [[SCALAR_RECUR_INIT]], [[SCALAR_PH]] ], [ [[REC_NEXT:%.*]], [[FOR_BODY]] ]
 ; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
 ; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P]], i64 [[IPROM]]
 ; CHECK-NEXT:    [[REC_NEXT]] = load i16, i16* [[B]], align 2
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    store i16 [[SCALAR_RECUR]], i16* [[B]], align 4
 ; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
 ; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END]], [[LOOP9:!llvm.loop !.*]]
 ; CHECK:       if.end:
-; CHECK-NEXT:    [[REC_LCSSA:%.*]] = phi i16 [ [[SCALAR_RECUR]], [[FOR_COND]] ], [ 10, [[FOR_BODY]] ], [ [[VECTOR_RECUR_EXTRACT_FOR_PHI]], [[MIDDLE_BLOCK]] ]
+; CHECK-NEXT:    [[REC_LCSSA:%.*]] = phi i16 [ [[SCALAR_RECUR]], [[FOR_COND]] ], [ 10, [[FOR_BODY]] ]
 ; CHECK-NEXT:    ret i16 [[REC_LCSSA]]
 ;
 entry:
   br label %for.cond
 
 for.cond:
   %i = phi i32 [ 0, %entry ], [ %inc, %for.body ]
   %rec = phi i16 [0, %entry], [ %rec.next, %for.body ]

llvm/test/Transforms/LoopVectorize/interleaved-accesses.ll

 ; CHECK-NEXT:    [[TMP3:%.*]] = and i64 [[INDEX]], 9223372036854775804
 ; CHECK-NEXT:    [[TMP4:%.*]] = getelementptr inbounds i32, i32* [[B:%.*]], i64 [[TMP3]]
 ; CHECK-NEXT:    [[TMP5:%.*]] = bitcast i32* [[TMP4]] to <4 x i32>*
 ; CHECK-NEXT:    store <4 x i32> [[TMP2]], <4 x i32>* [[TMP5]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 4
 ; CHECK-NEXT:    [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 508
 ; CHECK-NEXT:    br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP12:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 1016, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.cond.cleanup:
 ; CHECK-NEXT:    ret void
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
 ; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV]]
 ; CHECK-NEXT:    [[TMP:%.*]] = load i32, i32* [[ARRAYIDX]], align 4
 ; CHECK-NEXT:    [[MUL:%.*]] = shl nsw i32 [[TMP]], 1
 ; CHECK-NEXT:    [[TMP1:%.*]] = lshr exact i64 [[INDVARS_IV]], 1
 ; CHECK-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[TMP1]]
 ; CHECK-NEXT:    store i32 [[MUL]], i32* [[ARRAYIDX2]], align 4
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 2
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV]], 1022
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_COND_CLEANUP]], [[LOOP13:!llvm.loop !.*]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_COND_CLEANUP:%.*]], [[LOOP13:!llvm.loop !.*]]
 ;
 entry:
   br label %for.body
 
 for.cond.cleanup:                                 ; preds = %for.body
   ret void
 
 for.body:                                         ; preds = %for.body, %entry

 ; CHECK-NEXT:    [[TMP9:%.*]] = and i64 [[INDEX]], 9223372036854775804
 ; CHECK-NEXT:    [[TMP10:%.*]] = getelementptr inbounds i32, i32* [[B:%.*]], i64 [[TMP9]]
 ; CHECK-NEXT:    [[TMP11:%.*]] = bitcast i32* [[TMP10]] to <4 x i32>*
 ; CHECK-NEXT:    store <4 x i32> [[TMP8]], <4 x i32>* [[TMP11]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 4
 ; CHECK-NEXT:    [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP14:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    br i1 false, label [[FOR_COND_CLEANUP:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[IND_END]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.cond.cleanup:
 ; CHECK-NEXT:    ret void
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
 ; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV]]
 ; CHECK-NEXT:    [[TMP:%.*]] = load i32, i32* [[ARRAYIDX]], align 4
 ; CHECK-NEXT:    [[MUL:%.*]] = shl nsw i32 [[TMP]], 1
 ; CHECK-NEXT:    [[TMP1:%.*]] = lshr exact i64 [[INDVARS_IV]], 1
 ; CHECK-NEXT:    [[ARRAYIDX2:%.*]] = getelementptr inbounds i32, i32* [[B]], i64 [[TMP1]]
 ; CHECK-NEXT:    store i32 [[MUL]], i32* [[ARRAYIDX2]], align 4
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 2
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i64 [[INDVARS_IV_NEXT]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_COND_CLEANUP]], [[LOOP15:!llvm.loop !.*]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_COND_CLEANUP:%.*]], [[LOOP15:!llvm.loop !.*]]
 ;
 entry:
   br label %for.body
 
 for.cond.cleanup:                                 ; preds = %for.body
   ret void
 
 for.body:                                         ; preds = %for.body, %entry

 ; CHECK-NEXT:    [[TMP17:%.*]] = extractelement <8 x i32> [[WIDE_VEC]], i32 4
 ; CHECK-NEXT:    store i32 [[TMP17]], i32* [[TMP12]], align 4
 ; CHECK-NEXT:    [[TMP18:%.*]] = extractelement <8 x i32> [[WIDE_VEC]], i32 6
 ; CHECK-NEXT:    store i32 [[TMP18]], i32* [[TMP13]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 4
 ; CHECK-NEXT:    [[TMP19:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP24:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
 ; CHECK-NEXT:    [[P_I_X:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 0
 ; CHECK-NEXT:    [[P_I_Y:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 1
 ; CHECK-NEXT:    store i32 [[Z]], i32* [[P_I_X]], align 4
 ; CHECK-NEXT:    store i32 [[Z]], i32* [[P_I_Y]], align 4
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i64 [[I]], 1
 ; CHECK-NEXT:    [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
-; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], [[LOOP25:!llvm.loop !.*]]
+; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END:%.*]], [[LOOP25:!llvm.loop !.*]]
 ; CHECK:       for.end:
 ; CHECK-NEXT:    ret void
 ;
 entry:
   br label %for.body
 
 for.body:
   %i = phi i64 [ %i.next, %for.body ], [ 0, %entry ]

 ; CHECK-NEXT:    [[TMP18:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP18]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP26:!llvm.loop !.*]]
 ; CHECK:       middle.block:
 ; CHECK-NEXT:    [[RDX_SHUF:%.*]] = shufflevector <4 x i32> [[TMP17]], <4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
 ; CHECK-NEXT:    [[BIN_RDX:%.*]] = add <4 x i32> [[TMP17]], [[RDX_SHUF]]
 ; CHECK-NEXT:    [[RDX_SHUF3:%.*]] = shufflevector <4 x i32> [[BIN_RDX]], <4 x i32> poison, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
 ; CHECK-NEXT:    [[BIN_RDX4:%.*]] = add <4 x i32> [[BIN_RDX]], [[RDX_SHUF3]]
 ; CHECK-NEXT:    [[TMP19:%.*]] = extractelement <4 x i32> [[BIN_RDX4]], i32 0
-; CHECK-NEXT:    br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP19]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ]
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
 ; CHECK-NEXT:    [[S:%.*]] = phi i32 [ [[TMP21:%.*]], [[FOR_BODY]] ], [ [[BC_MERGE_RDX]], [[SCALAR_PH]] ]
 ; CHECK-NEXT:    [[P_I_X:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 0
 ; CHECK-NEXT:    [[P_I_Y:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 1
 ; CHECK-NEXT:    [[TMP20:%.*]] = load i32, i32* [[P_I_X]], align 4
 ; CHECK-NEXT:    store i32 [[TMP20]], i32* [[P_I_Y]], align 4
 ; CHECK-NEXT:    [[TMP21]] = add nsw i32 [[TMP20]], [[S]]
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i64 [[I]], 1
 ; CHECK-NEXT:    [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
-; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], [[LOOP27:!llvm.loop !.*]]
+; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END:%.*]], [[LOOP27:!llvm.loop !.*]]
 ; CHECK:       for.end:
-; CHECK-NEXT:    [[TMP22:%.*]] = phi i32 [ [[TMP21]], [[FOR_BODY]] ], [ [[TMP19]], [[MIDDLE_BLOCK]] ]
-; CHECK-NEXT:    ret i32 [[TMP22]]
+; CHECK-NEXT:    ret i32 [[TMP21]]
 ;
 entry:
   br label %for.body
 
 for.body:
   %i = phi i64 [ %i.next, %for.body ], [ 0, %entry ]
   %s = phi i32 [ %2, %for.body ], [ 0, %entry ]
   %p_i.x = getelementptr inbounds %pair.i32, %pair.i32* %p, i64 %i, i32 0

 ; CHECK-NEXT:    [[TMP18:%.*]] = extractelement <8 x i32> [[WIDE_VEC]], i32 4
 ; CHECK-NEXT:    store i32 [[TMP18]], i32* [[TMP13]], align 4
 ; CHECK-NEXT:    [[TMP19:%.*]] = extractelement <8 x i32> [[WIDE_VEC]], i32 6
 ; CHECK-NEXT:    store i32 [[TMP19]], i32* [[TMP14]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 4
 ; CHECK-NEXT:    [[TMP20:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP20]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP28:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
 ; CHECK-NEXT:    [[P_I_X:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 0
 ; CHECK-NEXT:    [[P_I_MINUS_1_X:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 -1, i32 0
 ; CHECK-NEXT:    [[P_I_Y:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 1
 ; CHECK-NEXT:    store i32 [[Z]], i32* [[P_I_X]], align 4
 ; CHECK-NEXT:    [[TMP21:%.*]] = load i32, i32* [[P_I_MINUS_1_X]], align 4
 ; CHECK-NEXT:    store i32 [[TMP21]], i32* [[P_I_Y]], align 4
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i64 [[I]], 1
 ; CHECK-NEXT:    [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
-; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], [[LOOP29:!llvm.loop !.*]]
+; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END:%.*]], [[LOOP29:!llvm.loop !.*]]
 ; CHECK:       for.end:
 ; CHECK-NEXT:    ret void
 ;
 entry:
   br label %for.body
 
 for.body:
   %i = phi i64 [ %i.next, %for.body ], [ 0, %entry ]

 ; CHECK-NEXT:    [[TMP21:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP21]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP30:!llvm.loop !.*]]
 ; CHECK:       middle.block:
 ; CHECK-NEXT:    [[RDX_SHUF:%.*]] = shufflevector <4 x i32> [[TMP20]], <4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
 ; CHECK-NEXT:    [[BIN_RDX:%.*]] = add <4 x i32> [[TMP20]], [[RDX_SHUF]]
 ; CHECK-NEXT:    [[RDX_SHUF3:%.*]] = shufflevector <4 x i32> [[BIN_RDX]], <4 x i32> poison, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
 ; CHECK-NEXT:    [[BIN_RDX4:%.*]] = add <4 x i32> [[BIN_RDX]], [[RDX_SHUF3]]
 ; CHECK-NEXT:    [[TMP22:%.*]] = extractelement <4 x i32> [[BIN_RDX4]], i32 0
-; CHECK-NEXT:    br i1 false, label [[FOR_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP22]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ]
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[I:%.*]] = phi i64 [ [[I_NEXT:%.*]], [[FOR_BODY]] ], [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ]
 ; CHECK-NEXT:    [[S:%.*]] = phi i32 [ [[TMP25:%.*]], [[FOR_BODY]] ], [ [[BC_MERGE_RDX]], [[SCALAR_PH]] ]
 ; CHECK-NEXT:    [[I_PLUS_1:%.*]] = add nuw nsw i64 [[I]], 1
 ; CHECK-NEXT:    [[P_I_X:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 0
 ; CHECK-NEXT:    [[P_I_Y:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I]], i32 1
 ; CHECK-NEXT:    [[P_I_PLUS_1_Y:%.*]] = getelementptr inbounds [[PAIR_I32]], %pair.i32* [[P]], i64 [[I_PLUS_1]], i32 1
 ; CHECK-NEXT:    [[TMP23:%.*]] = load i32, i32* [[P_I_X]], align 4
 ; CHECK-NEXT:    store i32 [[TMP23]], i32* [[P_I_PLUS_1_Y]], align 4
 ; CHECK-NEXT:    [[TMP24:%.*]] = load i32, i32* [[P_I_Y]], align 4
 ; CHECK-NEXT:    [[TMP25]] = add nsw i32 [[TMP24]], [[S]]
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i64 [[I]], 1
 ; CHECK-NEXT:    [[COND:%.*]] = icmp slt i64 [[I_NEXT]], [[N]]
-; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END]], [[LOOP31:!llvm.loop !.*]]
+; CHECK-NEXT:    br i1 [[COND]], label [[FOR_BODY]], label [[FOR_END:%.*]], [[LOOP31:!llvm.loop !.*]]
 ; CHECK:       for.end:
-; CHECK-NEXT:    [[TMP26:%.*]] = phi i32 [ [[TMP25]], [[FOR_BODY]] ], [ [[TMP22]], [[MIDDLE_BLOCK]] ]
-; CHECK-NEXT:    ret i32 [[TMP26]]
+; CHECK-NEXT:    ret i32 [[TMP25]]
 ;
 entry:
   br label %for.body
 
 for.body:
   %i = phi i64 [ %i.next, %for.body ], [ 0, %entry ]
   %s = phi i32 [ %2, %for.body ], [ 0, %entry ]
   %i_plus_1 = add nuw nsw i64 %i, 1

llvm/test/Transforms/LoopVectorize/loop-form.ll

 ; CHECK-NEXT:    [[TMP8:%.*]] = getelementptr inbounds i16, i16* [[TMP7]], i32 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = bitcast i16* [[TMP8]] to <2 x i16>*
 ; CHECK-NEXT:    store <2 x i16> zeroinitializer, <2 x i16>* [[TMP9]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i32 [[INDEX]], 2
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <2 x i32> [[VEC_IND]], <i32 2, i32 2>
 ; CHECK-NEXT:    [[TMP10:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP4:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i32 [[TMP1]], [[N_VEC]]
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[IF_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
 ; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P]], i64 [[IPROM]]
 ; CHECK-NEXT:    store i16 0, i16* [[B]], align 4
 ; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
 ; CHECK-NEXT:    br label [[FOR_COND]], [[LOOP5:!llvm.loop !.*]]
 ; CHECK:       if.end:
 ; CHECK-NEXT:    ret void

 ; CHECK-NEXT:    [[TMP9:%.*]] = getelementptr inbounds i16, i16* [[TMP8]], i32 0
 ; CHECK-NEXT:    [[TMP10:%.*]] = bitcast i16* [[TMP9]] to <2 x i16>*
 ; CHECK-NEXT:    store <2 x i16> zeroinitializer, <2 x i16>* [[TMP10]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i32 [[INDEX]], 2
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <2 x i32> [[VEC_IND]], <i32 2, i32 2>
 ; CHECK-NEXT:    [[TMP11:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP6:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i32 [[TMP2]], [[N_VEC]]
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[IF_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
 ; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P]], i64 [[IPROM]]
 ; CHECK-NEXT:    store i16 0, i16* [[B]], align 4
 ; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
 ; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END]], [[LOOP7:!llvm.loop !.*]]
 ; CHECK:       if.end:

 ; CHECK-NEXT:    [[TMP9:%.*]] = getelementptr inbounds i16, i16* [[TMP8]], i32 0
 ; CHECK-NEXT:    [[TMP10:%.*]] = bitcast i16* [[TMP9]] to <2 x i16>*
 ; CHECK-NEXT:    store <2 x i16> zeroinitializer, <2 x i16>* [[TMP10]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i32 [[INDEX]], 2
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <2 x i32> [[VEC_IND]], <i32 2, i32 2>
 ; CHECK-NEXT:    [[TMP11:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP8:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i32 [[TMP2]], [[N_VEC]]
-; CHECK-NEXT:    [[IND_ESCAPE:%.*]] = sub i32 [[N_VEC]], 1
-; CHECK-NEXT:    [[IND_ESCAPE1:%.*]] = sub i32 [[N_VEC]], 1
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[IF_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
 ; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P]], i64 [[IPROM]]
 ; CHECK-NEXT:    store i16 0, i16* [[B]], align 4
 ; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
 ; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END]], [[LOOP9:!llvm.loop !.*]]
 ; CHECK:       if.end:
-; CHECK-NEXT:    [[I_LCSSA:%.*]] = phi i32 [ [[I]], [[FOR_BODY]] ], [ [[I]], [[FOR_COND]] ], [ [[IND_ESCAPE1]], [[MIDDLE_BLOCK]] ]
+; CHECK-NEXT:    [[I_LCSSA:%.*]] = phi i32 [ [[I]], [[FOR_BODY]] ], [ [[I]], [[FOR_COND]] ]
 ; CHECK-NEXT:    ret i32 [[I_LCSSA]]
 ;
 ; TAILFOLD-LABEL: @multiple_unique_exit2(
 ; TAILFOLD-NEXT:  entry:
 ; TAILFOLD-NEXT:    br label [[FOR_COND:%.*]]
 ; TAILFOLD:       for.cond:
 ; TAILFOLD-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; TAILFOLD-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N:%.*]]

 ; CHECK-NEXT:    [[TMP9:%.*]] = getelementptr inbounds i16, i16* [[TMP8]], i32 0
 ; CHECK-NEXT:    [[TMP10:%.*]] = bitcast i16* [[TMP9]] to <2 x i16>*
 ; CHECK-NEXT:    store <2 x i16> zeroinitializer, <2 x i16>* [[TMP10]], align 4
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i32 [[INDEX]], 2
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <2 x i32> [[VEC_IND]], <i32 2, i32 2>
 ; CHECK-NEXT:    [[TMP11:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]]
 ; CHECK-NEXT:    br i1 [[TMP11]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP10:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i32 [[TMP2]], [[N_VEC]]
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[IF_END:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ [[N_VEC]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N]]
-; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
 ; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P]], i64 [[IPROM]]
 ; CHECK-NEXT:    store i16 0, i16* [[B]], align 4
 ; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
 ; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
 ; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END]], [[LOOP11:!llvm.loop !.*]]
 ; CHECK:       if.end:
-; CHECK-NEXT:    [[EXIT:%.*]] = phi i32 [ 0, [[FOR_COND]] ], [ 1, [[FOR_BODY]] ], [ 0, [[MIDDLE_BLOCK]] ]
+; CHECK-NEXT:    [[EXIT:%.*]] = phi i32 [ 0, [[FOR_COND]] ], [ 1, [[FOR_BODY]] ]
 ; CHECK-NEXT:    ret i32 [[EXIT]]
 ;
 ; TAILFOLD-LABEL: @multiple_unique_exit3(
 ; TAILFOLD-NEXT:  entry:
 ; TAILFOLD-NEXT:    br label [[FOR_COND:%.*]]
 ; TAILFOLD:       for.cond:
 ; TAILFOLD-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
 ; TAILFOLD-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N:%.*]]

 
 if.end:
   ret i32 0
 
 if.end2:
   ret i32 1
 }
 
+; LCSSA, common value each exit
+define i32 @multiple_exit_blocks2(i16* %p, i32 %n) {
+; CHECK-LABEL: @multiple_exit_blocks2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[FOR_COND:%.*]]
+; CHECK:       for.cond:
+; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
+; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N:%.*]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
+; CHECK:       for.body:
+; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
+; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P:%.*]], i64 [[IPROM]]
+; CHECK-NEXT:    store i16 0, i16* [[B]], align 4
+; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
+; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
+; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END2:%.*]]
+; CHECK:       if.end:
+; CHECK-NEXT:    [[I_LCSSA:%.*]] = phi i32 [ [[I]], [[FOR_COND]] ]
+; CHECK-NEXT:    ret i32 [[I_LCSSA]]
+; CHECK:       if.end2:
+; CHECK-NEXT:    [[I_LCSSA1:%.*]] = phi i32 [ [[I]], [[FOR_BODY]] ]
+; CHECK-NEXT:    ret i32 [[I_LCSSA1]]
+;
+; TAILFOLD-LABEL: @multiple_exit_blocks2(
+; TAILFOLD-NEXT:  entry:
+; TAILFOLD-NEXT:    br label [[FOR_COND:%.*]]
+; TAILFOLD:       for.cond:
+; TAILFOLD-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
+; TAILFOLD-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N:%.*]]
+; TAILFOLD-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
+; TAILFOLD:       for.body:
+; TAILFOLD-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
+; TAILFOLD-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P:%.*]], i64 [[IPROM]]
+; TAILFOLD-NEXT:    store i16 0, i16* [[B]], align 4
+; TAILFOLD-NEXT:    [[INC]] = add nsw i32 [[I]], 1
+; TAILFOLD-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
+; TAILFOLD-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END2:%.*]]
+; TAILFOLD:       if.end:
+; TAILFOLD-NEXT:    [[I_LCSSA:%.*]] = phi i32 [ [[I]], [[FOR_COND]] ]
+; TAILFOLD-NEXT:    ret i32 [[I_LCSSA]]
+; TAILFOLD:       if.end2:
+; TAILFOLD-NEXT:    [[I_LCSSA1:%.*]] = phi i32 [ [[I]], [[FOR_BODY]] ]
+; TAILFOLD-NEXT:    ret i32 [[I_LCSSA1]]
+;
+entry:
+  br label %for.cond
+
+for.cond:
+  %i = phi i32 [ 0, %entry ], [ %inc, %for.body ]
+  %cmp = icmp slt i32 %i, %n
+  br i1 %cmp, label %for.body, label %if.end
+
+for.body:
+  %iprom = sext i32 %i to i64
+  %b = getelementptr inbounds i16, i16* %p, i64 %iprom
+  store i16 0, i16* %b, align 4
+  %inc = add nsw i32 %i, 1
+  %cmp2 = icmp slt i32 %i, 2096
+  br i1 %cmp2, label %for.cond, label %if.end2
+
+if.end:
+  ret i32 %i
+
+if.end2:
+  ret i32 %i
+}
+
+; LCSSA, distinct value each exit
+define i32 @multiple_exit_blocks3(i16* %p, i32 %n) {
+; CHECK-LABEL: @multiple_exit_blocks3(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[FOR_COND:%.*]]
+; CHECK:       for.cond:
+; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
+; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N:%.*]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
+; CHECK:       for.body:
+; CHECK-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
+; CHECK-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P:%.*]], i64 [[IPROM]]
+; CHECK-NEXT:    store i16 0, i16* [[B]], align 4
+; CHECK-NEXT:    [[INC]] = add nsw i32 [[I]], 1
+; CHECK-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
+; CHECK-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END2:%.*]]
+; CHECK:       if.end:
+; CHECK-NEXT:    [[I_LCSSA:%.*]] = phi i32 [ [[I]], [[FOR_COND]] ]
+; CHECK-NEXT:    ret i32 [[I_LCSSA]]
+; CHECK:       if.end2:
+; CHECK-NEXT:    [[INC_LCSSA:%.*]] = phi i32 [ [[INC]], [[FOR_BODY]] ]
+; CHECK-NEXT:    ret i32 [[INC_LCSSA]]
+;
+; TAILFOLD-LABEL: @multiple_exit_blocks3(
+; TAILFOLD-NEXT:  entry:
+; TAILFOLD-NEXT:    br label [[FOR_COND:%.*]]
+; TAILFOLD:       for.cond:
+; TAILFOLD-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY:%.*]] ]
+; TAILFOLD-NEXT:    [[CMP:%.*]] = icmp slt i32 [[I]], [[N:%.*]]
+; TAILFOLD-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[IF_END:%.*]]
+; TAILFOLD:       for.body:
+; TAILFOLD-NEXT:    [[IPROM:%.*]] = sext i32 [[I]] to i64
+; TAILFOLD-NEXT:    [[B:%.*]] = getelementptr inbounds i16, i16* [[P:%.*]], i64 [[IPROM]]
+; TAILFOLD-NEXT:    store i16 0, i16* [[B]], align 4
+; TAILFOLD-NEXT:    [[INC]] = add nsw i32 [[I]], 1
+; TAILFOLD-NEXT:    [[CMP2:%.*]] = icmp slt i32 [[I]], 2096
+; TAILFOLD-NEXT:    br i1 [[CMP2]], label [[FOR_COND]], label [[IF_END2:%.*]]
+; TAILFOLD:       if.end:
+; TAILFOLD-NEXT:    [[I_LCSSA:%.*]] = phi i32 [ [[I]], [[FOR_COND]] ]
+; TAILFOLD-NEXT:    ret i32 [[I_LCSSA]]
+; TAILFOLD:       if.end2:
+; TAILFOLD-NEXT:    [[INC_LCSSA:%.*]] = phi i32 [ [[INC]], [[FOR_BODY]] ]
+; TAILFOLD-NEXT:    ret i32 [[INC_LCSSA]]
+;
+entry:
+  br label %for.cond
+
+for.cond:
+  %i = phi i32 [ 0, %entry ], [ %inc, %for.body ]
+  %cmp = icmp slt i32 %i, %n
+  br i1 %cmp, label %for.body, label %if.end
+
+for.body:
+  %iprom = sext i32 %i to i64
+  %b = getelementptr inbounds i16, i16* %p, i64 %iprom
+  store i16 0, i16* %b, align 4
+  %inc = add nsw i32 %i, 1
+  %cmp2 = icmp slt i32 %i, 2096
+  br i1 %cmp2, label %for.cond, label %if.end2
+
+if.end:
+  ret i32 %i
+
+if.end2:
+  ret i32 %inc
+}
+
+
 ; unique exit case but with a switch as two edges between the same pair of
 ; blocks is an often missed edge case
 define i32 @multiple_exit_switch(i16* %p, i32 %n) {
 ; CHECK-LABEL: @multiple_exit_switch(
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_COND]] ]

 ; CHECK-NEXT:    store float 1.000000e+01, float* [[TMP9]], align 4
 ; CHECK-NEXT:    br label [[PRED_STORE_CONTINUE2]]
 ; CHECK:       pred.store.continue2:
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 2
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 2, i64 2>
 ; CHECK-NEXT:    [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], 200
 ; CHECK-NEXT:    br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP12:!llvm.loop !.*]]
 ; CHECK:       middle.block:
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 201, 200
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 200, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    br label [[LOOP_HEADER:%.*]]
 ; CHECK:       loop.header:
 ; CHECK-NEXT:    [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP_LATCH:%.*]] ]
 ; CHECK-NEXT:    [[GEP:%.*]] = getelementptr float, float* [[ADDR]], i64 [[IV]]
 ; CHECK-NEXT:    [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV]], 200
-; CHECK-NEXT:    br i1 [[EXITCOND_NOT]], label [[EXIT]], label [[LOOP_BODY:%.*]]
+; CHECK-NEXT:    br i1 [[EXITCOND_NOT]], label [[EXIT:%.*]], label [[LOOP_BODY:%.*]]
 ; CHECK:       loop.body:
 ; CHECK-NEXT:    [[TMP11:%.*]] = load float, float* [[GEP]], align 4
 ; CHECK-NEXT:    [[PRED:%.*]] = fcmp oeq float [[TMP11]], 0.000000e+00
 ; CHECK-NEXT:    br i1 [[PRED]], label [[LOOP_LATCH]], label [[THEN:%.*]]
 ; CHECK:       then:
 ; CHECK-NEXT:    store float 1.000000e+01, float* [[GEP]], align 4
 ; CHECK-NEXT:    br label [[LOOP_LATCH]]
 ; CHECK:       loop.latch:

 ; CHECK-NEXT:    [[INDEX_NEXT]] = add i64 [[INDEX]], 2
 ; CHECK-NEXT:    [[VEC_IND_NEXT]] = add <2 x i64> [[VEC_IND]], <i64 2, i64 2>
 ; CHECK-NEXT:    [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 200
 ; CHECK-NEXT:    br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], [[LOOP14:!llvm.loop !.*]]
 ; CHECK:       middle.block:
 ; CHECK-NEXT:    [[RDX_SHUF:%.*]] = shufflevector <2 x i32> [[TMP5]], <2 x i32> poison, <2 x i32> <i32 1, i32 undef>
 ; CHECK-NEXT:    [[BIN_RDX:%.*]] = add <2 x i32> [[TMP5]], [[RDX_SHUF]]
 ; CHECK-NEXT:    [[TMP7:%.*]] = extractelement <2 x i32> [[BIN_RDX]], i32 0
-; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 201, 200
-; CHECK-NEXT:    br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
+; CHECK-NEXT:    br label [[SCALAR_PH]]
 ; CHECK:       scalar.ph:
 ; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 200, [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY:%.*]] ]
 ; CHECK-NEXT:    [[BC_MERGE_RDX:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ]
 ; CHECK-NEXT:    br label [[LOOP_HEADER:%.*]]
 ; CHECK:       loop.header:
 ; CHECK-NEXT:    [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP_LATCH:%.*]] ]
 ; CHECK-NEXT:    [[ACCUM:%.*]] = phi i32 [ [[BC_MERGE_RDX]], [[SCALAR_PH]] ], [ [[ACCUM_NEXT:%.*]], [[LOOP_LATCH]] ]
 ; CHECK-NEXT:    [[GEP:%.*]] = getelementptr i32, i32* [[ADDR]], i64 [[IV]]
 ; CHECK-NEXT:    [[EXITCOND_NOT:%.*]] = icmp eq i64 [[IV]], 200
-; CHECK-NEXT:    br i1 [[EXITCOND_NOT]], label [[EXIT]], label [[LOOP_LATCH]]
+; CHECK-NEXT:    br i1 [[EXITCOND_NOT]], label [[EXIT:%.*]], label [[LOOP_LATCH]]
 ; CHECK:       loop.latch:
 ; CHECK-NEXT:    [[TMP8:%.*]] = load i32, i32* [[GEP]], align 4
 ; CHECK-NEXT:    [[ACCUM_NEXT]] = add i32 [[ACCUM]], [[TMP8]]
 ; CHECK-NEXT:    [[IV_NEXT]] = add nuw nsw i64 [[IV]], 1
 ; CHECK-NEXT:    [[EXITCOND2_NOT:%.*]] = icmp eq i64 [[IV]], 400
 ; CHECK-NEXT:    br i1 [[EXITCOND2_NOT]], label [[EXIT]], label [[LOOP_HEADER]], [[LOOP15:!llvm.loop !.*]]
 ; CHECK:       exit:
-; CHECK-NEXT:    [[LCSSA:%.*]] = phi i32 [ 0, [[LOOP_HEADER]] ], [ [[ACCUM_NEXT]], [[LOOP_LATCH]] ], [ [[TMP7]], [[MIDDLE_BLOCK]] ]
+; CHECK-NEXT:    [[LCSSA:%.*]] = phi i32 [ 0, [[LOOP_HEADER]] ], [ [[ACCUM_NEXT]], [[LOOP_LATCH]] ]
 ; CHECK-NEXT:    ret i32 [[LCSSA]]
 ;
 ; TAILFOLD-LABEL: @me_reduction(
 ; TAILFOLD-NEXT:  entry:
 ; TAILFOLD-NEXT:    br label [[LOOP_HEADER:%.*]]
 ; TAILFOLD:       loop.header:
 ; TAILFOLD-NEXT:    [[IV:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[IV_NEXT:%.*]], [[LOOP_LATCH:%.*]] ]
 ; TAILFOLD-NEXT:    [[ACCUM:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[ACCUM_NEXT:%.*]], [[LOOP_LATCH]] ]