diff --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -3488,8 +3488,19 @@
       Optional<TargetTransformInfo::ShuffleKind> ShuffleKind =
           isShuffle(VL, Mask);
       if (ShuffleKind.hasValue()) {
-        InstructionCost Cost =
-            TTI->getShuffleCost(ShuffleKind.getValue(), VecTy, Mask);
+        InstructionCost Cost = 0;
+
+        if (*ShuffleKind != TargetTransformInfo::SK_PermuteSingleSrc)
+          Cost += TTI->getShuffleCost(ShuffleKind.getValue(), VecTy, Mask);
+
+        unsigned NumOfParts = TTI->getNumberOfParts(VecTy);
+        // Compute the number of number of elements per vector register for
+        // VecTy. If that is not possible, because the number of parts for VecTy
+        // is unknown, use the maximum value for unsigned.
+        unsigned EltsPerVector =
+            NumOfParts ? VecTy->getNumElements() / NumOfParts : -1;
+        unsigned Idx = 0;
+
         for (auto *V : VL) {
           // If all users of instruction are going to be vectorized and this
           // instruction itself is not going to be vectorized, consider this
@@ -3502,6 +3513,26 @@
             Cost -= TTI->getVectorInstrCost(Instruction::ExtractElement, VecTy,
                                             IO->getZExtValue());
           }
+          if (*ShuffleKind == TargetTransformInfo::SK_PermuteSingleSrc) {
+            if ((Idx > 0 && (Idx + 1) % EltsPerVector == 0) ||
+                Idx + 1 == VL.size()) {
+              SmallVector<int> Mask;
+              unsigned StartIdx =
+                  Idx - std::min(EltsPerVector, VecTy->getNumElements()) + 1;
+              Optional<TargetTransformInfo::ShuffleKind> ShuffleKind =
+                  isShuffle(
+                      {&VL[StartIdx],
+                       std::min(EltsPerVector, unsigned(VL.size() - StartIdx))},
+                      Mask);
+              Cost += TTI->getShuffleCost(
+                  ShuffleKind.getValue(),
+                  FixedVectorType::get(
+                      VecTy->getElementType(),
+                      std::min(EltsPerVector, VecTy->getNumElements())),
+                  Mask);
+            }
+          }
+          ++Idx;
         }
         return ReuseShuffleCost + Cost;
       }