diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h b/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
index f80379b..8f6a73d 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
@@ -535,13 +535,13 @@
   /// Returns true if the per-lane cost of VectorizationFactor A is lower than
   /// that of B.
   bool isMoreProfitable(const VectorizationFactor &A,
-                        const VectorizationFactor &B) const;
+                        const VectorizationFactor &B, bool HasTail) const;
 
   /// Returns true if the per-lane cost of VectorizationFactor A is lower than
   /// that of B in the context of vectorizing a loop with known \p MaxTripCount.
   bool isMoreProfitable(const VectorizationFactor &A,
                         const VectorizationFactor &B,
-                        const unsigned MaxTripCount) const;
+                        const unsigned MaxTripCount, bool HasTail) const;
 
   /// Determines if we have the infrastructure to vectorize the loop and its
   /// epilogue, assuming the main loop is vectorized by \p VF.
diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index a4d546f..249e8fc 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -4256,9 +4256,10 @@
   return EstimatedVF;
 }
 
-bool LoopVectorizationPlanner::isMoreProfitable(
-    const VectorizationFactor &A, const VectorizationFactor &B,
-    const unsigned MaxTripCount) const {
+bool LoopVectorizationPlanner::isMoreProfitable(const VectorizationFactor &A,
+                                                const VectorizationFactor &B,
+                                                const unsigned MaxTripCount,
+                                                bool HasTail) const {
   InstructionCost CostA = A.Cost;
   InstructionCost CostB = B.Cost;
 
@@ -4296,9 +4297,9 @@
   if (!MaxTripCount)
     return CmpFn(CostA * EstimatedWidthB, CostB * EstimatedWidthA);
 
-  auto GetCostForTC = [MaxTripCount, this](unsigned VF,
-                                           InstructionCost VectorCost,
-                                           InstructionCost ScalarCost) {
+  auto GetCostForTC = [MaxTripCount, HasTail](unsigned VF,
+                                              InstructionCost VectorCost,
+                                              InstructionCost ScalarCost) {
     // If the trip count is a known (possibly small) constant, the trip count
     // will be rounded up to an integer number of iterations under
     // FoldTailByMasking. The total cost in that case will be
@@ -4307,9 +4308,10 @@
     // some extra overheads, but for the purpose of comparing the costs of
     // different VFs we can use this to compare the total loop-body cost
     // expected after vectorization.
-    if (CM.foldTailByMasking())
-      return VectorCost * divideCeil(MaxTripCount, VF);
-    return VectorCost * (MaxTripCount / VF) + ScalarCost * (MaxTripCount % VF);
+    if (HasTail)
+      return VectorCost * (MaxTripCount / VF) +
+             ScalarCost * (MaxTripCount % VF);
+    return VectorCost * divideCeil(MaxTripCount, VF);
   };
 
   auto RTCostA = GetCostForTC(EstimatedWidthA, CostA, A.ScalarCost);
@@ -4317,10 +4319,12 @@
   return CmpFn(RTCostA, RTCostB);
 }
 
-bool LoopVectorizationPlanner::isMoreProfitable(
-    const VectorizationFactor &A, const VectorizationFactor &B) const {
+bool LoopVectorizationPlanner::isMoreProfitable(const VectorizationFactor &A,
+                                                const VectorizationFactor &B,
+                                                bool HasTail) const {
   const unsigned MaxTripCount = PSE.getSmallConstantMaxTripCount();
-  return LoopVectorizationPlanner::isMoreProfitable(A, B, MaxTripCount);
+  return LoopVectorizationPlanner::isMoreProfitable(A, B, MaxTripCount,
+                                                    HasTail);
 }
 
 void LoopVectorizationPlanner::emitInvalidCostRemarks(
@@ -4609,7 +4613,7 @@
         continue;
       }
 
-      if (isMoreProfitable(Candidate, ChosenFactor))
+      if (isMoreProfitable(Candidate, ChosenFactor, P->hasScalarTail()))
         ChosenFactor = Candidate;
     }
   }
@@ -4623,7 +4627,8 @@
   }
 
   LLVM_DEBUG(if (ForceVectorization && !ChosenFactor.Width.isScalar() &&
-                 !isMoreProfitable(ChosenFactor, ScalarCost)) dbgs()
+                 !isMoreProfitable(ChosenFactor, ScalarCost,
+                                   !CM.foldTailByMasking())) dbgs()
              << "LV: Vectorization seems to be not beneficial, "
              << "but was forced by a user.\n");
   return ChosenFactor;
@@ -4789,7 +4794,7 @@
     }
 
     if (Result.Width.isScalar() ||
-        isMoreProfitable(NextVF, Result, MaxTripCount))
+        isMoreProfitable(NextVF, Result, MaxTripCount, !CM.foldTailByMasking()))
       Result = NextVF;
   }
 
@@ -7768,11 +7773,11 @@
 
       InstructionCost Cost = cost(*P, VF);
       VectorizationFactor CurrentFactor(VF, Cost, ScalarCost);
-      if (isMoreProfitable(CurrentFactor, BestFactor))
+      if (isMoreProfitable(CurrentFactor, BestFactor, P->hasScalarTail()))
         BestFactor = CurrentFactor;
 
       // If profitable add it to ProfitableVF list.
-      if (isMoreProfitable(CurrentFactor, ScalarFactor))
+      if (isMoreProfitable(CurrentFactor, ScalarFactor, P->hasScalarTail()))
         ProfitableVFs.push_back(CurrentFactor);
     }
   }
diff --git a/llvm/lib/Transforms/Vectorize/VPlan.h b/llvm/lib/Transforms/Vectorize/VPlan.h
index 7cdcb24..60c9b1c 100644
--- a/llvm/lib/Transforms/Vectorize/VPlan.h
+++ b/llvm/lib/Transforms/Vectorize/VPlan.h
@@ -3790,6 +3790,13 @@
   bool hasEarlyExit() const {
     return ExitBlocks.size() > 1 || ExitBlocks[0]->getNumPredecessors() > 1;
   }
+
+  /// Returns true if the scalar tail may execute after the vector loop. Note
+  /// that this relies on unneeded branches to the scalar tail loop being
+  /// removed.
+  bool hasScalarTail() const {
+    return getScalarPreheader()->getNumPredecessors() != 0;
+  }
 };
 
 #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)